Maybe there should be two-dimensions: Speed and Accuracy.

You could tell your researchers you want them to come up with a bow with longer range and in six months they will do it.

Or you could tell your researchers you want them to come up with a better bow and in four months they will do it, but you don't know if it will be more accurate, better range, easier to reload, faster to produce, etc. 

Or you could tell your researchers you want them to come up with someway to improve your military and in three months they will do it, but you don't know if it will come up with better armor, better bows, etc.

Or you could tell your researchers you want them to come up with... something and in two months they will do it, but you have no clue what they will give you. Could be military, could be agriculture, could be contruction, etc.

Some might think it is better to allow your researchers freedom to come up with as many new ideas as possible, while others will want to focus on the core techs to make them better ASAP.

It might be fun to let the researchs go wild for a while and then use those techs as a guide for your empire.

"Three of my first five techs were from Agriculture. My people are natural farmers."

Make it a mix on several levels

Magic tech should be random but you shouldn't be able to get better spells until youve researched their pre-reqs (Like AoW:SM)

Regular tech weve seen screnes for and it seems likely that youll have a linear tree up to a point, afterword you can likely continue investing for "+1+ style benefits. That or random (I say this because they already said thier will be infinite techs)

Beta's only a few weeks away. soon we can get a better idea.

Balancing it would be hard. It would have to be a tough choice between trying to get as many techs as fast as possible or concentrating on getting the techs needed most right away.

But I think it is realistic. After all, you tell 100 scientists, 'Come up with someone.' at least one of them will be able to have a breakthrough relatively fast. But if you tell the same 100 scientists, 'Come up with a longer range for my bows.' some won't have any idea how to do that and won't be useful.

You could grow so much food that you could have legions of poorly trained, poorly equiped soldiers that could flood into an area and take it over.

You could also trade food to starving empires in exchange for protection, and military technology.

This is the kind of thinking that shows how awesome a random system can be. Some people argue against it on the panicky basis that you will never get anything good, but even if you do have a bad run, it simply makes you come up with new strat. I dunno, i like the idea of thinking differently each game. I guess doing the same thing over and over has its value too...

Thank you Frogboy you  have made my day. I just knew you and I thought a lot alike. :{)

Agreed don't let reloaders cheat make them play the game to the end or start over no reload get another chance ability.