Oh man, this topic... Forget a wall of text I'm going to tower of babel this shit.
The thing is for most genres multiplayer can be a good value add. But it all depends how it's implemented. But implemented right it's the easiest way to add the much sort over "replayability" factor. Of course, it's also easy to abuse so it's a mixed back. But it also has a lot of secondary benefits.
Take FPS for example. Singleplayer campaigns are becoming rarer and rarer because so little of player's time is spent on them. Players play through the story once and love it or hate it they then move onto multiplayer and spend 3 times as long there on average. It's further worth noting that is brought down by players who spend a short amount of time with the game. The long-term player base that makes up the persistent communities for these games spend the overwhelming majority of their time with the game in multiplayer.
Part of the attraction in normal human competitiveness. Part of it is a balance issue. There is a sweet spot for AI difficulty in shooters. You can have the AI instakill you with flawless headshots every time you pass by a knothole in the wall they can see you through if you want. This does not make for a positive player experience so most AI is intentionally nerfed so the player feels they can compete. Then there's the flip side where the AI shoots like storm troopers and the player mows through them effortlessly. This power fantasy is appealing to some but for others, it becomes quickly boring due to the lack of challenge. Multiplayer eliminates this balancing act entirely. Other humans are on the same scale as you and are bound by the same rulesets by default. From there if it's a fast-paced game or slow-paced is simply how much health you give players and you don't risk players feeling cheated on either end.
But this doesn't mean there's not a market for singleplayer experiences in shooters. Wolfenstein and Doom had major success with the singleplayer games recently. Doom is a prime example that making a good multiplayer game and a good singleplayer game have very different requirements as the much as the singleplayer was praised the multiplayer was heavily criticised. But you have the same issue, players play through the story, the majority don't even complete it. They might come back and restart it a few times over the following months. But then they leave. This is fine from an initial sales point of view but the problem is most won't come back for an expansion. You will get collectors buying the expansion but not playing it. This also sounds fine from a sales point of view but you lose word of mouth about the expansion and it's not going to bring in new players to buy the base game as a result so it generally better to release a whole new game to get there attention again... Which can lead to franchise fatigue.
As we established the majority (like 80%) of players don't complete good length singleplayer campaigns, to begin with. If you keep the campaign short enough that they do they feel like they didn't get their monies worth. Ironically the illusion of value is based on getting to the end being too much for most people. Then you bring out another game for $60, most existing players haven't completed the first one... There's a sense it's going to be more of the same. You are basically trying to monetise monotony to them.
By contrast when there is a sequel to the multiplayer you are selling a new dynamic. Human players stay the same but now there are new maps, new weapons, new strategies. Multiplayer players are perpetually excited for new content. This also includes the controversial trickle feed microtransaction content. Something that if you include in a singleplayer game you can expect to be review bombed into oblivion. Who wants to buy vanity items for singleplayer games? No one else is going to see you wearing them. Buying things for the stats? Most people are old enough to remember cheat codes and difficulty settings, the notion of selling power to a player in a singleplayer game is an insult. If your game is so grindy you need to buy advantages to speed it up, it's not well balanced singleplayer experience to begin with.
Strategy/tactics games have the most difficult niche. You have the extreme end of the AI never been competitive enough for experienced players without feeling like the AI is cheating. Playing with human leads to more compelling diplomatic interactions, more balanced competition and is pretty much universally a more positive experience, particularly if players can create their own races and role-play them. Short form this is a win-win, so quick turn around RTS games, much like FPS, thrive on the quality of their multiplayer game. Long form however you run into massive issues in turn based games waiting for your opponents and having other players stick around long enough to complete the game. People have lives. Getting a group together to play a game that can take multiple days to play is very hard. Then you have the issue of players being removed from play early on. Plus once you fall behind you can rarely catch up. You get painfully awkward situations were either wipe a player out or drag them along forcing them to play knowing they can never win.
There is always the option to play with random strangers. There are a lot of problems here too. While you can happily wipe a stranger out of the game because there's no accountability, that's a double-edged sword because they can just up and leave on a whim. You can invest hours getting a game started just to have other players abandon the game. Co-ordinating strangers to keep a game going for several days is very hard. Play by post is a solution in turn based games for some of this but it takes waiting for other players to the extreme and really takes you out of the game. In this day and age you can get your gaming fix immediately from AI so waiting a week for a player to take a turn is madness.
But what about other genres? RPG's and Adventures? There is a huge gulf between an MMO and a traditional singleplayer RPG/action adventure. MMO's have the appeal of community interaction, stranger danger and shared growth. But the trade-off for all this is almost universally any depth in storytelling and singleplayer significance. That's not to say you can't reach a balance here, but developers inevitably won't because once you start an MMO your community drives income and your community demands community content, leaving lone wolf players in the cold.
Singleplayer games in these genres are facing a lot of the same issues as FPS SP campaigns. The only difference is RPG's take a lot longer to develop so franchise fatigue is traditionally slower. Or it was... The golden era of gaming means consumers are inundated with more games in their chosen genre than they can realistically give time to. So they aren't just competing with themselves when it comes to wearing players out. One of the big loopholes that have been found is to have a short singleplayer story the player can knock over in a couple of hours if they want, but also have a huge sandbox world full other content they explore for ages creating replayability separate from the story. This also lets you scratch the grind itch more naturally. Trouble is everyone is making that those now too.
But there's still a huge market for those games so why the move to multiplayer? Because MMO's and competitive frag-fests aren't the only alternatives and when you look at the statistics you find that pretty much all consumers will play it. As much as the media pushes the opposing narrative, gaming is inherently a social activity. Even sticking to singleplayer games there is a drive for player to want to share their single player experience with other gamers in their social circle. Appart from meeting and talking about it, we encourage others to watch us play (some even making a living out of that) and people will even compete still comparing high scores, or in their absence, who has the highest level/stats, who can complete the game fastest etc. Yet there is still a drive to bring other players into your singleplayer gameworld to interact with. Even if just for a little while. In a given household/family there is also the matter of sharing time. Got one TV in the loungeroom, who gets to play on it? Why are you spending time playing singleplayer instead of spending time with your spouse/kids? I mean ideally, you could do both...
To that end, optional co-op is almost always a universal win-win. If you can be playing your single player game happily for the most the day and then when your spouse gets home if they can just seamlessly join you and share the experience that's great. If they can wander off and do their own thing and then meet back up with you that's extremely optimal. You tick almost every box (you just need optional arena they can visit and compete in and you've got the full package). A lot of developers are realising this. The only reason every game doesn't ship with co-op is it's harder to develop. Multiplayer network programming is its own separate skills package. Then there's extra design work making sure the game mechanics work in multiplayer and all around more money and time spent on delivering the same length game. But it will be received better because there is no downside to it and for the players it matters to, it matters a lot to. So generally if a developer is capable, and it makes sense to do so, they will try to include at least co-op multiplayer if they can.
At the same time, you have a lot of publishers looking at the statistics and drawing tangental conclusions. The strong move towards multiplayer from a publisher perspective is about money. Multiplayer means more players if one person is a group buys the game they are going to encourage others to buy it to play with them. Multiplayer games are self-marketing in addition to being easier to monetise, review better, suffer less franchise fatigue and more importantly clearly target the larger market demographic. So the publisher doesn't see as much value in singleplayer game campaigns... But they are wrong.
Even if you look at MMO's the statistics also show that most players don't want to be dealing with hostile randoms on the internet. Not at all, this is why the market only seems large enough to maintain a handful of successful MMO's at a time. Also why MMO's that restrict PVP are a lot more successful than open PVP clusterfucks. Fans for those exist but they are in the minority. There is a slightly larger group that like a sense of danger but the thing is they want downtime too and more importantly they want things to be fair. Getting ganked by 5 players twice your level might be okay as a one-off. No one is going to enjoy it happening everytime they try to do anything. The other problem MMO's have is players can go to other genres for that fair danger fix. There are hundreds of small-scale server-based multiplayer PVP games in multiple genres they can go to and even find one with more or less the right balance for their personal taste. More reason for publishers to make non-MMO games multiplayer yes?
Yes but it's not a good reason to dump the single player. As I said that group wants two things, a fair playing field and downtime. Jumping into a death match with a bunch of experienced players that keep sniping you while you are trying to learn the controls is no fun. This is why a lot of people won't play the FPS franchise titles that don't come with a singleplayer campaign. They want to learn how to play the game and feel accomplished in to before moving to multiplayer. The singleplayer campaign also lets them take a break from the multiplayer and play at their own pace every once in a while. Even though most players will spend very little time in the single player, it is very important to maximising your player base to have both. Most developers understand this, convincing the committees running the publisher is another matter.
But take for example the big sandbox games which are such a huge market at the moment. You add large-scale multiplayer capability and you get GTA5 online. Stupidly successful. So much so the publisher can't even comprehend why it's so successful and keeps making a bunch of stupid decisions leading to torrents of hate from fans and onlookers. In fact, it is perfectly reasonable to say that on the whole, GTA5 online is generally quite strongly hated, even by people who play it. So why the success? Branding obviously but also the lack of alternatives at the time. It scratches several itches. A giant sandbox is fun to fuck around in, it's even more fun with someone else to fuck around with. You've got the sense the danger. But the bigs one?
Mutliplayer gives your grind more value. All the time you spent exploring, gathering better equipment, fighting and honing your real life skills in addition to raising up your stats. It's good for single player progression... But it's even better when you can show it off to other people in the game. Being able to work together with friends to speed up this progression for both of you is just as appealing. Take a game like Zelda Breath of the Wild. As satisfying as it is as a single player game, who is going to say no to letting a friend visiting with their Nintendo Switch loading into your game as a Shadow Link and engaging in activities like racing between points across the land, or working together co-operatively in the trial of the sword or even having an arena match with one another? It's added gameplay that makes all those little tools and advancements you've accumulated to make your character more capable even more valuable thanks to the shared experience. Your accomplishments are worth more when another player gives them context. Co-op based shrine puzzles would just be four swords worth of gravy. Speaking Nintendo look at the pokemon games. Even right at the start you have an amazingly singleplayer focused series of games that was given so much more depth by the ability to link up with friends and trade monsters you've trained and pointlessly battle them. In 60 hours of playtime maybe an hour was spent actually doing those things but it made the other 59 hours matter so much more than they did on their own.
So what the problem with GTA5 online? Well, basically not understanding the middle ground the game actively encourages the worst possible behaviour from players you randomly encounter. As established the people who are into that are in the minority. Then there are the publisher dumbs. Seeing what a money spinner it was they got greedy. The grind in the game is beyond reasonable. What is the point in spending 10,000 hours grinding (and that's not an exaggeration) to accomplish certain things when another player can dump $$$ down and get it instantly and rub it in your face? Then there's the issue of content focus. Further development into the game was directed towards making even more grind and monitisable assets. There has been no more singleplayer gameplay added to the game since launch. The first DLC's that added interesting new missions to the multiplayer have been abandoned in favour of making stuff players pretty much have to buy to stay competitive instead. So the game is now a cesspool and is largely despised even by initial fans (That's ignoring other publisher stupidity like banning modding too).
By contrast, take similar games like just cause or far cry adding optional co-op on top of their singleplayer experience (an important distinction because making a co-op centric game impairs the development of the singleplayer aspects which are so important) and you get a universally positive response. Then contrast the open servers vs the private servers of open sandbox survival games. The open servers are cesspools like GTA online and are largely despised except for the minority. Private server lets you not only play co-operate with your friends and compete with them on your own terms, but you can also setup competing teams/tribes/clans across the server creating that sense of danger without the accompanying douchebaggery you get with it from randoms in open servers. So there is a very positive response there.
Now let's discuss the elephant in the room. Fallout 76. They have clearly added all tools necessary of open survival server cesspool madness. As much as I seem to be speaking down on that element, this is smart. See players who like that are in the minority, but they still exist and more importantly, they are very vocal and loud. Encouraging that element is amazing free advertising and why should they have things they enjoy? That said it was loosely implied that Fallout 76 also has friend filtering. If you can play online on an empty server by yourself, then occasionally have your friends and only your friends jump in an join you, cooperatively and competitively that's an amazing proposal. Even the option to build something in private then invite the masses in to do battle with you on your home turf sounds pretty great doesn't it? Basically, the success of Fallout 76 will come down to 2 things. The strength of the singleplayer content and the flexibility of server filtering for the multiplayer content. Well, I say that... It will be successful no matter what because of branding. I mean it's the next Fallout title. Ongoing success, however, will depend on those 2 things.
So yeah it's no surprise we are seeing more multiplayer capable games at E3. Publishers see the $$$ signs and developers see the potential for enhanced player interaction. It doesn't inherently devalue the singleplayer experiences. When done right it just adds to them... Both ways. Greater than the sum of their parts and all that.
So what about Stardock? Well, they kind of exist in this weird bubble. On the one hand, they have third parties making games that flop like demigod. But we know that was largely a timing issue with the rise of free to play games in that genre... Not to mention the launch travesty caused by rampant piracy... Sin of Solar empire was an amazing game that had it's multiplayer community wounded at launch. You had a long-form game that a couple of hours in it would end to the dreaded Out of Sync error (That was the same issue that kept The Guild 2 from being anything more than a niche title). Who wants to invest the time in organising long-term multiplayer match for it just end like that? I'm sure it got better but I've been afraid to start a Sins game in recent times because I just don't know if it was every truly comprehensibly fixed.
Galactic Civ 3 is missing all kinds of love when to comes to the multiplayer game. The genre has infrequent multiplayer activity as it is. You are going to see substantially less multiplayer activity if you don't work on the quality of life improvements to address that. Offworld Trading Company is incredibly cutthroat and exponential growth nature of the power creep means you get what is known as runaway leaders. From the other players, perspective things spiral out of control so fast once certain thresholds have been crossed that if you blink you'll miss the events that unfolded that lead to you losing. The game is also really bad at telegraphing information about such things to new players so even in my own group of competently intelligent new players, after a couple of multiplayer matches in which one player hits on a winning strategy, it's back to singleplayer for several hours for the entire group to play with the mechanics more to figure out whats going on and how to deal with it. Rince and repeat several times and then queue more singleplayer playtime once you all have the gameplay down to practice your reaction times between multiplayer play sessions.
I don't have access to your metrics, but given your niche and how things have unfolded with Stardock's games, I would be amazed if those metrics didn't show an utterly negligible amount of multiplayer activity in your multiplayer capable titles. Even the ones built for multiplayer weren't primed to encourage it. Even if they were things like Galactic Civ would only see a modest bump due to the difficulty of multiplayer communities co-ordinating for the necessary timeframes. Pretty much the optimal multiplayer scenario for a 4X games means they are on a scale where pacing and inconsistent presence doesn't matter as much. It almost looked like you were heading that way when you were experimenting with the mega/multiverse leaderboard thing for GC2?
There is nothing to say there aren't clever ways to incorporate multiplayer into something like GC3 that would see constant use by the community. Those things just have to be limited to short encounters that complements the single-player game. Something like allowing empires that finished the game based on the turn limit end condition (victorious or not) to compete with one another in a short vicious conflict to see which ascended empire is greater when galaxies collide. You spend the whole single player game fine turning your civilisation for the ultimate battle after the single player game is over and building/expansion no longer matters. The turn limit restriction being the balancing factor. Being able to take that empire you built to war with your friend's empire adds value the time investment of building it after the singleplayer game is over. This does, of course, require players to play their singleplayer game while connected to an online server to prevent cheating (bonus is this acts as a discouragement to piracy as you'd need a licenced account to play the game that way)
Another example is asynchronous multiplayer encounters during a singleplayer game. Again you'd need to play the singleplayer game while connected to an online server. Between turns, the game could then search the games of other online players and find appropriate matchups for crossover events. Three examples based on one concept: An anomaly opens a spacial rift between dimensions, time seems to move at a different pace between dimension but more interesting com signals reveal alien empires from other worlds exist beyond the rift.
Varaition 1: Enter interdimensional space piracy. A player is given an option to send a fleet through the rift, pausing their singleplayer game for a moment to enter the other player's dimension where they can raid on their map for a fixed number of turns before the rift closing pulls them back to their own dimension.
Variation 2: The player can select a fleet to offer as a bounty hunter to the other player's dimension. Other players and AI can select a target and bid on who the fleet will attack when it comes through the rift. The player's game is paused while they take control of their fleet in the other players match and get paid based on how much damage they do to their selected target in the fixed number of turns before the rift closes snaping their fleet back to their own dimension with the payment.
Variations 3: Like variant 2 the player can offer a fleet for players/ai in the other game to bid on but this time as a mercenary. The player receives immediate payment and their fleet will leave the game. The winner of the bid in the other game will gain control of the fleet for X number of turns before the rift closes and the fleet snaps back to the other dimension of origin. On the next turn in their game, the original player will receive whatever remains of their fleet. This variant doesn't require players to play with each other in each other's games, but it does allow players to show off their custom ship designs in a fulfilling interactive way.
Bonus Variant: Instead of conflict lets talk trade. Certain strategic resources not being on a map in any abundance is a common problem. When the player finds an anomaly the matchmaker could look for other players with an abundance of a resource the player doesn't have access to. You could then send trade ships through the rift and require them to be defended on both sides to maintain the trade route. Or imagine a variant where the time the rift is open is based on real-time rather than turns. The number of trade vessels each player manages to send through the rift before it closes will determine how many turns the resource will be available for after it closes. Encourage quick frantic play for some minutes before returning to careful planning.
Now provided you give players to opt out of the types of encounters they don't like in their games you'll I think you'd be surprised how many players opt into this kind of multiplayer interaction interrupting their single player game. Especially if you also enable friend only filtering for specific types of encounters.
TLDR:
There is a lot more to 'multiplayer' than the simplified media representation of what it generally entails. If you are a PC gamer and you don't play some form of online multiplayer you literally are in the 1%. But that doesn't mean you are geared towards the massive online clusterfuck style action games publishers seem to the think is the promised land and even if you are that doesn't mean a comprehensive single player experience doesn't have value to you. Furthermore it is often true those singleplayer experiences have more value if paired with a multiplayer experience and vice versa. There is absolutely a market for a competitive version of star control, but more importantly, I can tell you right now a lot of people would be seriously pissed if you did that instead of making sure it had a comprehensive well fleshed out singleplayer experience. I can also tell you no one would be upset if you added an optional co-op mode to that comprehensive singleplayer game after it was completed. Also, your multiplayer support for Galactic Civ is shit. 
