Stopped by IRC today,

https://forums.elementalgame.com/408950 

Actually that population mechanic sounds great.

I would describe it for fluff as the following:

Diminishing returns occur because it takes more manpower to accomplish the same amount of work. For example, the impact one researcher has on a society (the first researcher) is greater than the second researcher. Throwing more people at a problem does not have a linear affect. Cooperation and time management decrease efficiency overall. Trying to coordinate 1000 scientists is harder than coordinating 100 and as a result, the "science per person" drops.

This change will also make population a more important resource.

 If per empire, conquering of cities is going to mess it up somehow, and something that will need testing.

One additonal idea: I'd suggest adding a lvl 4 and lvl 5 school-type building that gives +% research to cities you are in, kinda like merchant and market for wealth.  I think your idea will create a need for such buildings.

Frogboy, are you going to extend this idea to troops? Maybe the 2nd, 3rd and four training tier take up that many people per person in the unit.

ie,

int popUsed = trainingLevel*unitSize;

That doesn't make all that much sense. Even assuming for the sake of argument that absolutely no education takes place in the new towns you are building in the barren world (which is a pretty weird assumption given how they are slowly built up to be shining centres of learning and civilization), the available expertise will be more a function of population size than of previously hired expertise. Sure, when you hire expertise the pool of available expertise grows smaller, but when you are building your empire in the wastelands and attracting people your pool of available expertise grows much faster from increased population than it is decreased by hiring.

The logical implications of your implementation is that it is damn easy to find 3 monks and archivists amongst a population of the first 6 people attracted to your new hole in the ground while you'll need to attract at least 465 to find 30 of them, and 5050 to find 100 of them.

Do it if you think it improves game balance and believe players will find destroying large parts of the cities they conquer in order to be allowed to recruit units or build buildings of their own fun, but don't, for the love of god, do it for realism. How about keeping population cost fixed and the RP total taken to some suitably chosen power in the interval 0.4-0.7? Sure, people will still spam studies, but they do take time to produce, and this will slow down the research growth significantly.

 EDIT: I see you've already addressed the destruction issue in the post you posted while I was writing. Good for you.

-----

Of course, what Elemental really needs is to have (at least) two manpower resources - the unwashed masses and the educated class, growing separately and the educated class' growth dependant primarily on the civilization level of a settlement (size, different types of buildings present) with a tiny base immigration bonus per turn. Any cowpoke can pick up a spear - making an educated man takes effort.