I disagree with research trees semi-randomly developing.  If anything, make this an option.  Here's why:  I like to play Rebellion against 9 computers set to Vicious.  It's only possible if you know which technologies will be available.  For Tech its important to get a great economy including Trade and Pervasive Economy early.  For Vasari, starbases and their upgrades are critical.  For Advent, there's a couple ways to go, but culture spamming is the easiest.  If I had to muddle my way into finding critical subjects, I'd be slaughtered. 

I think allowing repeat research is the way to avoid a " cookie-cutter research path".

One of my favorite custom mods for Rebellion sets the maximum level for most research subjects to 100.  It really allows for specialization.   If you want 100 levels of shields, they get expensive, but they are nice.  One siege frigate powerful enough to blow up a planet is possible, unless the opponent has researched enough ground defenses to render them useless.  I've seen light frigates more powerful than capital ships.  And I like that level of specialization.  The AI doesn't know how to cope with the changes, so they can be beaten on any level (with the exception of an occasional Advent stalemate condition).  But it's hilarious falling back to an overwhelming fleet of carriers, until they jump in to a system maxed out with starbases to the point no more will fit in the gravity well.  Then hearing a friend shouting and cussing across the room.  Having enough research depth that even though everything is unlocked, research can still progress is a great way to add variability to the game.

I'm not a fan of the tier system.  It just doesn't logically make sense to me, to be at point where research stops until you research how to research.  It would make more sense if the tiers unlocked as soon as you researched half the prior tier and had the proper amount of research labs (I guess if you absolutely hated the prior tier you could research general studies to bypass it.) 

I also prefer setting the research labs to only required one Civilian slot instead of 4.  This allows players to choose to specialize instead of expand earlier, especially in a small map or asteroid heavy environment.

I realize the game is in the brainstorming and experimental phase, so I apologize for sounding like I'm shooting down ideas before they get explored.  Ideally, when a new idea comes out, I'd like a check box to enable or disable it and see which version I enjoy better.  Absent of the ability to test both ways, I feel I have to play them out in my mind and anticipate what I would like better.  In this case, I know that if the map and players dictate I choose one type of strategy over another, I don't want to be hamstrung by not being able to research it.  

I'm a huge fan of the original system but I do think adjustments or improvements could be made to it.  When it comes to upgrading  anything it will be based on use of the object the more you use the that certain object the faster you can research it.(You will be able to research all techs but I think some techs should take longer so that this idea works better)

Example: Upgrading lasers for TEC will be based on frigates you have so like cobalt frigate uses only lasers so for each of those frigates you make its a .10x for researching anything having to do with lasers. For lasers shirken corvettes use multiple different types of weapons So they would increase  the laser tech by .02x so on so forth for other frigates. Upgrading hull, health, and shield will be determined by damage taken and health/shield restored.  

For the civilian tree the more orbital structures built the faster you can research other orbital structures

the more enemy ships you have enter your gravity well the faster you can research defensive buildings and/or research certain techs that make you build  faster and/or be able to see a phase lane or 2 away. 

Basically what I'm saying is cause and effect the more you do something or see something the better you get at it is the idea im going for nothing to difficult.  So I think this would be great for a sprawling empire or for a defensive empire it would assist in both ways it makes you  better in whatever field you would need help in.

Not saying this has to be added too. But maybe if you have so much of a object it automatically gives you the next level of that objects tech to you because you mastered that tech. This could be a breaking feature that allows the player to kind of research that one tech without having to build  a civilian or military research building of that level to research  that item.

I also preferred the system with the separated research trees and I am not sure about the tier system tbh. If one needs to research the tech tier for each military and civilian separately, why not implement two different research labs instead to begin with? It is better for scouting. While I like that the team wants to focus on the lore and removed lasers for the TEC, I liked the weapon diversity and the tactical/strategic diversity that came with it (as stated by someone in this thread before). In the end what matters to me however, is how many upgrades there will be and how they will affect ones strategy. What also could be an option is that there are "paths" in research and when deciding for one, the other one will be locked for the rest of the match. (As either one of them may eventually lead to a strong tech that may be game changing for example.)

Another idea is to retain the basic research system the game already has, but introduce a chance that any particular research project will fail.  For the sake of argument, let’s make it 10%.

So your research failed.  Maybe you get one chance to try again, spending the same points all over.  This time, if you fail again, you can’t try again until you’ve acquired the next highest tech tier.

Then you can try again, and this time you get a 99% chance of success.  But there would still be a slight chance that you could never acquire the tech you were after.

There might be ways to deal with this problem; various work-arounds.  But the basic idea is leave the system the way it is but introduce a random element that can ruin your day.

Maybe you could agree to spend extra money on research in order to reduce your risk of failure, or do the opposite and rush the research, but increase the chance it will bomb and you will get ‘tech blocked’.