How much would  it change the game if  research of a tech "failed" or was delayed?  Must we always be successful?

So what you're saying is that you'd love a game where you don't get an iron mine at the start of the game, and be completely overrun by your neigbour that does?

Research is about pacing, and giving the game phases. If you see that your neighbour has iron and you do not, you know that about 1/4 of the way through the game you are going to be in serious trouble. You now have options to research different paths to prepare for that time, and attempt to minimise it by attempting to move to the next phase as quickly as possible. In a similar vein, if you are the one with and early iron deposit, you would be looking to maximise that advantage by researching mining before, say, magic.

Research, or some level of advancement, is a foundational concept of 4X strategy games. It adds another resource that you need to manage (research points or the like) to create strategic advantages. That said, some of the advancements seem rather artificial (like unlocking hero / exploration). I can see the reasoning and process behind why those limits are there, but it doesn't seem to be part of the world, just an artificial limit to enhance strategy. I don't know whether it makes the game less 'fun' for me, but it does break the sense of immersion. Ideally game mechanics should support and extend the feeling of being part of the game and, once discovered, reveal a deeper understanding of how he world works as well as the game.

To me the whole reason these games are so damn addictive is because "in just one more turn"  we'll discover a shard, or found a new city, or produce a great new unit, or unlock a cool new technology via research.  When done right it's fantastic, you can certainly have a fun game without it, but I'd rather have it than not. 

However, if I had to pick between tech research and spell research I'd vote for spell research over tech research at least for this game, since I'm all for this actually being as close to a successor of MoM as we can reasonably hope for.

My main problem with research so far has been it just takes too damn long down certain paths, and some of the both early and late research items seem to serve no useful purpose.

I think every resource should have its own tech tree with like 10 levels. Then you could specialize your armies with enchantments from crystals or stainless steel swords from iron or whatever. It would at least be a choice to make and it would not be the same from game to game.

I like the research tree. Although it could use some tinkering, to cut it out entirely is unwise IMO.

I like research - I'm always playing very research oriented so I always got better troops than my enemies.

The only problem I got with research is that Warfare obsoletes my units pretty darn quickly and upgrading them is not really cheap. And that sucks, I often end up disbanding a bunch of spearman cause their wages are so damn high and upgrading them is also damn costy - so I need to hire new units... :-/

I like these ideas. This is the kind of thing that will make each replay feel like quite a different game which I think would be a good thing.

Pacing can be done with buildings as well as technology, or a combination of both.

I dislike research when it is used purely as a crutch, and posted as much about the civics tree earlier. The developers have already stated they don't intend to make any 'major' changes, which means removing the tech tree entirely probably won't happen. Personally I have no issues with it, in fact it is probably considerably easier to balance pacing using build times and numbers of cities, than with a research mechanic that could potentially be chosen in the 'wrong' order along with a number of other random factors such as maximum potential population and growth.

This is my biggest qualm with the research mechanic currently: If you fail to research civics at the right times, you gimp your empire. Additionally ignoring production and just choosing the higher population city is almost always the optimal route.

If they are insistant on using the tech model, research should be based on more than just population.

The research in this game could use a bit of tweaking (I'd prefer easier to research nodes that granted smaller advantages), I'm of the opinion that research is a great way to simulate progression of society from a dark-age post-apocalypse into a newer renaissance.

I like your ideas, cephalo.

Researching technology or unlocking through buildings can both work in a game like this. Some kind of hybrid would be great I think, as they can balance each other.
Researching allows for improvement in performance of units, equipment, and buildings, without necessarily requiring changes in infrastructure. But it also allows for some weird things like being able to train fully armor clad knights in a newly build settlement.
Unlocking through buildings doesn't have this, and it's easier to pace I think. But it also allows for beeline tactics that are pretty cheesy and ruin games for me.

What I really don't like about research in FE is that the hated 'unlocking game features through research' mechanic from WoM is still there.

Why won't certain heroes join me until I have researched a certain tech, and then still require gold to join? Give them some interesting reasons for joining or refusing to. One could require that my sovereign be at least 1 level higher than him, so he knows I'm worthy to follow. One could require that I have a city with at least 1000 people in it, so he knows I'm trying to rebuild civilization. A few should require me doing a quest for them. A few might even require a certain tech being researched, not because it unlocks recruiting them but because they are interested in it. So one mage type hero might not join until I have researched a certain part of the magic tree, and another one might not join until I have researched Literacy because he always wanted to learn how to read but never got the opportunity to learn.

Why can't I train groups of more than 3 soldiers at the start of the game? And why can't I have more than 5 of these units together in one army? Decisions like this should be about costs and benefits, and necessity, not some arbitrary tech level.

Same thing with quests. I can understand the victory quest chain requiring research, but others not so much. There must be better ways of balancing when certain quests can be done. One quest could require a large army to beat; you can do this early but run the risk of having virtually no defences left, or wait untill later when the potential loss could be acceptable. Or it requires strong heroes; you could wait untill they are high level enough, or you could risk lower level heroes but accept that a few of them could die. Some quests could be impossible to beat untill you can deal with receiving massive amounts of fire damage. Some could require you to be able to dispell some devastating enemy debuffs. Or cast some debuffs of your own.

Makes you able to personalize your civilization, the most important thing for some of us.

I'll give you a compromise that I think would make the research tree less tiresome.

Slim it down quite a lot and make it more proportional to the nation and its resources as opposed to being a minimax numbers system as it is now.

Give random events and the real world searched libraries than can only occur when certain conditions are met that allow the use of unique technologies.

My biggest gripe with the tech tree is the universalism of some of its features that make no sense to be universal. For example I could have 4 Water shards but still learn to make huge flaming axes even if I had no fire shard. It would suit the aura of the game more if magic based techs and faction based techs were a response to play style rather than a chess game choice.

Giving a hypothetical. Imagine you're playing a Porcipine game and you've specialized in Water magic, as she comes across a lost library in the woods she explores the many tomes but finds she lacks the magical knowledge from some of the books to learn their secrets, but she stumbles upon a recipe for frost armor, given her mastery of water magic this is something she can grasp with insight and low and behold suddenly her army is equiped in the very beginnings of frost gear and a new short tech tree for frost gear, with pre-req techs, opens up for her and her alone on that map.

This would add immersive quality to the techs, making them dynamic and also stylistic. One of the key elements of magic world empires is that they have a unified theme. The Sith wear red and black and shoot lightning, the horde is spiky and toothed, egyptian mythology is incredibly visible in its use in stories, that is what is Elemental and still FE are missing is a uniqueness for the empires. Discoveries and themes that are theirs and theirs alone not because that is where they planned their skill tree, and eventually everyone else will catch up, but because of a combination of elements that lead to something special only they got this game.