i mostly focus on one tech tree at a time while ignore all  others only to have some start a war that i no tech for a good army pretty much same in Gal Civ 2, this would make people that like to blitz one tree to spread out more and not worry about needing 100% focus on one tree.

Picking and choosing is great when you [A] have a unified tree with all your options clearly displayed at every research event, or [B] have the entire game memorized. We don't have A, and B is actually weaker in Elemental because of random results. What I proposed in the last paragraph of the OP, however, IMO doesn't break the part of that which you like (and which I agree with) -- your choices at each breakthrough event are still important, and focusing on one area is still the only way to exceed in that area. It's not actually simultaneous research, it's just the ability to occasionally repurpose it. I'd even say that a penalty for doing this would be reasonable; e.g. if you pick something that came from a different category, you lose +50% or even double its research cost from your pool. (Or 1.62x. Mmm, phi.) Since there's an upward curve on research costs, that's still going to be less than the full cost of a breakthrough in your primary line.

Being a random chance and carrying a meaningful opportunity cost, I don't think this breaks the system in any way, or encourages laziness; whereas it does make the game a lot less punishing for new players, time-constrained multiplayer games where you'd rather spend your turn moving units on the map, people who just don't like that level of micromanagement, etc. It's more accessible and more forgiving, but not so forgiving that it's pointless to ever change paths. (It occurs to me that with infinite incremental end-game techs, this is less true. Picks used on those probably shouldn't count for the tech-level comparison, or you would eventually get everything just by researching one category. Mind you, it would take a lot longer.)

Again, remember that my core motivation here is that dictatorial science is a terribly unrealistic abstraction. Yes, funding generates ideas; but ideas also happen without funding, especially in areas where there hasn't been much progress. IMO that should be represented, and if you can come up with other ways to handle it, please chime in. I happen to like this one because it greatly improves the existing abstraction without significantly altering it, and has the accessibility benefits mentioned above. But maybe there are other options?

I actually logged on this morning to post something similar <.< lol 

What I was thinking though is something like you have a flat 10% chance every turn of researching something in each tree.  Whenever you build a building or recruit a unit that gives research points, you can invest those points into "actively researching" one of the tech trees.  As an example, say you have three research points and put them all into Warfare.  That could increase your chance of getting something in Warfare up to 13%, while the other four techs stay at 10%.  Since the player is "actively researching" warfare though the points could be cumulative, making it a 16% chance to learn something in warfare the next turn if something wasn't researched the previous turn, while keeping the others at 10%.  Once the current warfare tech is learned, the total could go back down to 13%. 

I think this could be fairly easily balanced by making higher tiers of research reduce the bonus % you get from research points.  Using the warfare example again, if you only invest 3 research points into it a T1 research would give you the full 3% per turn (so going up 3%, 6%, 9%, etc for each subsequent turn that the research it's discovered), but a T4 or T5 could reduce that bonus to something like 1% per turn (going up 1%, 2%, etc.).  This would only affect the bonus though, never dipping below the default 10% per turn.

Bump for 3-C.   Also edited title as I think the idea just above is much better than the original.