I would like for every faction to have its own Technologies (mostly) ... although I think it should not be in a particular "tree" which would imply 2-D relationships between tech progression, but more of a 3 or 4 dimensional "chain-mail" of technologies, where each technology has a tier (which alot of non-related techs sit upon) mainly to represent relative cost and relative effectiveness (or at least cost) ... and have a Tech on tier X have connections to several techs below it, and several techs above it.
The teching will be blind, although you start out with relatively equal chances to tech each of the tier 1 techs of Grouping A. Grouping A is one of the catagory groups of technologies, whether military, Agriculture, Architecture, ect.
Regardless of how much technology you set in order to research towards Group A, eventually you will research one of the techs, picked randomly. You get a pop-up that names the tech and describes the tech, with a few thematic and game=play applications mentioned in the description. Then, the percentage of the other Tier-1 techs stay the same, while the percent-chance of said researched tech is divided into several Tier-2 techs it could lead into. I do admit that in the behind-the-scenes, its a very simple and linear tree, but soon you could have, say, 5% chance to get one of a couple tier 10s, while you still have a 30% chance to get that tier 1 (in an extreme example) however probability will lend itself to a more natural progression of light tech -> heavy tech, although in some games you could get stuck with researching a string of high-end techs, but naturally they would take longer as well, and the probability is pretty low this would happen. However ... as I said you distribute tech between all of the Groupings, so lets say in Grouping D your extremely even and have tiers 1-3 completed (even if you don't entirely know this, but it could be memorized I suppose), while Grouping A has stringed together to get a tier 6 tech. You know there is a slim chance Grouping A is working on a Tier 7 tech, or tier 4, 5, 6 due to the raw-branching mechanism behind the scenes. Therefore, if you want to be more generalized, perhaps continue as normal, but if you feel you want expensive technologies, you know Grouping A has at least a chance you will get a high technology. Therefore you could devote all research to only Grouping A activities, in order to get whatever it is faster. Anyways, this logic is also very two-dimensional in make-up (therefore easy to code I think), although it provides alot of variety.
Add into that "hidden pathways" which only unlock if conditions are met (so suddenly a tier 2 tech is unlocked while your researching a tier 5 tech, after the tier 5 tech is researched, there will be a relatively good chance you could be researching that tier 2 tech, although you won't know until its researched ... although since its on such a low tier there is a high probability that it was a "hidden tech" which you finally completed the requirements for. Many random hidden techs, which provide a significant advantage/change to the game, will probably make organic development of Nation more fun, of course you can still choose how much percentage of full research you are assigning to each group.
I do not want Science and Economy to be directly related. I would rather research always be 100% ... but just deciding how many percentage points to go into each grouping. Also, this static percentage would be based upon your total research, which can improve with buildings and projects. The over-all level of Education will effect the 100% ... I have mentioned in a previous post, that Education in such a method should be finite maximum, to a maximum of 200% aka tech progression at double the raw research rate. I also mentioned that no more than 100 points could be dedicated to one group .. unless such extra points were added at half or a third of efficiency. (200 points to make 150 points).
In the end, I think it would make science more unique and separate from economy.