IMHO all future Stardock/IC games should come with a 64-bit build. We have seen over and over and over that 4X strategy games have a sprawling nature to their data set that is very difficult to contain within the virtual address space of a 32-bit process, at least as far as Stardock's implementations have shown. Couple that with some kind of high performance on-disk database for caching less often used objects/data, and you should form the technical foundation for the future of Stardock/Ironclad 4X strategy games. These two measures would in principle get rid of the minidumps due to out of memory once and for all.

Technical issues aside, I'd like to see a more immersive customizability and personality to the units you create and the planets. GalCiv 2 would let you see the planet closer up, and throw out quotes from a database representing how the citizens feel about living there. That was pretty cool, but that represents just one step towards the kind of character-building that I'd like to see.

Next, I'd like to see a veterancy system for ordinary ships and starbases, so that they can improve their stats somewhat as they have more battle experience. They won't be able to get new abilities or improve their stats as markedly as a capital ship, but there would be value in retaining your ships rather than throwing them to their deaths. This would have to be balanced so that a player without veteran ships would still have a fighting chance (given proper strategy) to take out the guy who has been training up a fleet of veteran ships all game.

For planets, there should be one-off morale boosts for certain events, that only last a limited time. Morale would slightly improve the rate of construction, population growth and commerce. You would notice a morale boost if a planet gets sieged but survives due to a hard-won battle in orbit with your defending forces or fleets. You might also get a morale boost in a friendly system where your flagship (the oldest capital ship currently in service) is currently stationed.

I also like the idea of evolving technology, hinted at by someone in an earlier post. Keep the tech tree conceptually as-is, but also factor in that ships naturally and irreparably deteriorate over the span of years. For example, it is only possible to retrofit a ship so far, because design decisions like the shape of the hull and the space in the interior of the ship will eventually reach hard limits, and changing these limits would be too expensive or impossible without compromising structural integrity.

This concept of evolving technology would manifest as a gradual decrease in effectiveness of long-serving capships, eventually necessitating the decommissioning of a ship. When a ship is decommissioned, its service record is evaluated, coming up with a numerical "score" that represents the ratio of engagements attempted versus engagements "won" (where enemy losses exceeded friendly). The higher the score, the more of the crew of the ship will be re-commissioned into a new ship, which will be stronger, more current and more powerful than the old ship. This will manifest itself in things like flight speed, weapon damage, hull points, and the effectiveness of activated abilities.

So if your flagship reaches a very high level and has only rarely had to retreat from a losing battle, while on the flip side it conquered many planets, it would be decommissioned retaining, say, 90% of the experience points of the original ship. A new ship of the same class would be created with stats even greater than the starting values of the original flagship, and it would keep 90% of the experience.

Decommissioning wouldn't become mandatory at any given time; you could in principle leave an ancient ship in service forever. But after a certain point, additional credits would start to be deducted to keep the ship in service, presumably because obsolete parts have to be special-ordered, and crews trained on new ships have to be trained also on the ancient ship. The effectiveness debuff would cap out around 75%; e.g., an initial hull integrity of 1000 would stop deteriorating at 250. But you can't just decommission a brand new ship, either; decommissioning wouldn't be possible at all until, say, 20% of the initial stats have been lost due to ship age. It's up to the player to decide whether to keep an old ship in service, or to turn it around (at a significant strategic risk due to its unavailability) as soon as decommissioning is possible. Maybe also throw in a researchable tech to improve the benefits granted to replacement ships when you decommission.

The upshot of the decommissioning process is that, near the end of a very long game, capital ships would be extremely powerful. Having gone through 6 or 7 generations of improvements, your highly experienced crew could come out of the factory with a new ship at level 7 (out of 10) but with 3 times the firepower and hull points as the original design at the same experience level. This rewards careful strategists who make sure that their capital ships survive engagements, and provides the maximum reward to those who make sure that they actually win every engagement involving their cap ships. Eventually, this effect becomes a self-fulfilling prophecy, helping to wind down painfully long games by providing the more skilled strategist with a fast, powerful behemoth of a capital ship that can roll in with a small escort fleet and conquer the enemy, regardless of the amount of resistance. But to get to that level, we're talking hours and hours and hours of play, accumulating momentum and field experience.

Aside from modelling the real life facts of design obsolescence, this system would also model the real life concept of operational knowledge, which is basically a whole set of competencies and rules that ship crewmen (in the modern Navy) accumulate over the course of their career. When the chain of command is unbroken, and senior officers are allowed to personally train green officers in the ways of the ship, the operational knowledge is preserved from generation to generation, even across new ships. If the ship is lost in combat, all that operational knowledge is lost. And we can assume that ships that do very poorly in combat won't have much valuable operational knowledge to share anyway, hence there would be a significant (50-60%) exp loss by decommissioning a ship that loses most of its engagements.

Just thought of something else: Maybe you can have the decommissioning process only apply for TEC, to add some flavor. For the Vasari, ships don't deteriorate, but you can invest materials / credits to unlock experience levels above 10, forming a sort of "super-elite" crew that can crank the ship stats into overdrive. Same result, immensely powerful capships, but without deterioration over time, instead bringing in a resource cost. Not sure what other unique system you could introduce for the Advent, but I'm sure someone could think of one.

Mechs!

I want the Sins 2 game to become more like :

Homeworld: Real 3D tactics, formation, ship movement, and lego like ship interconnectivity. Sins is so close to topping the king. If you combine Sins empire management with Homeworld's 3D, ship movements (no bumper boats) and formations. I would be in heaven.  

AI Wars: This game has an incredible arsonal of tactics. Sins 2 needs to become deeper with more complex ships.

Galactic Civilizations 2: Oh god, could you imaging Sins with the ability to customize ships like GC2

Gatatuious Space Battles: The ability to customize ships with modules and assign tactics

Dragon Age: Origins: Yes an RPG, but it had the ability to add complex tactics with the simplicity of setting rules with drop downs in Outlook. Imagine setting rules of engagement commands like "If enemy stucture is within 400km and health < 30% attach enemy stucture if gravity well contains no enemy capital ships"