why create a new mechanic to limit buildings by population when we have a perfectly good one already? ie, city level. i'd rather current game mechanics were better made use of and integrated than new ones added.
so now as well as checking back to level up my city, or every time a building is finished, i now need an update flashed across the screen every time a new slot is spawned as well? or will i need to keep checking myself? if it's anywhere near as fine-grained as brad implies, than i will have to do either fairly often.
if i literally get one slot/x number of citizens, that makes the number of slots between levels 4 and 5 vastly more than between 1 and 2, because the population requirements increase at an increasing rate. to me that implies either massive end game micro, or an even more limited and boring early game, because even more of the fun stuff gets locked away and hidden from the player at the start of the game. this end game micro will be more apparent if the population growth is still modelled with the same crude flat person/turn method, because we will still have a scenario where every settlement eventually ends up being huge when given enough time. far better to have populations level out at sustainable levels until circumstances change. that way we get a diverse range of settlements, with some places staying as small villages so i only ever have a few huge cities to manage.
far better to use the level system itself as a limiting mechanic, that way you neither get too many slots at high level, nor too few at low level (currently, you won't be able to get any before level 2) at the very least i would hope that if the specialist slots are implemented then they ditch the awful level-up bonuses, as they seem to be representing exactly the same thing. the level up bonuses are also arbitrary, obtuse and hard to keep track of. at the same time however, if they are removed then there is little reason to care about levelling up at all, except to qualify for the higher level buildings. which you could migrate that to the specialist slot mechanic as well (because it is representing the same thing), in which case why have levels at all?
in summary, either limit things by level, or ditch levels and add specialist slots. i don't see any advantage in having both. personally levels make more sense to me because they keep up with ballooning populations instead of getting ahead/behind them. if you want so say "you can only build two of the following in a level 3 settlement," then i'm sure there is an easier way to do it than creating a whole new mechanic.
personally i'm not that big on the idea of having demolish existing buildings to build more troops in any circumstance. nice bit of background and interesting theory in brad's post, but i don't see how it makes the game any more fun or balanced.
i realise that what i'm arguing for sounds like the unfun status quo, but the reason it's not fun is because of the method used to determine population growth and the fact that the production of that population is not related to how many of them there are. you CAN do that by saying "one merchant per 10 people," but it's much easier to do so by simply saying "merchant generates 0.05gildar per person." you then represent your increasing number of specialists by allowing this to be increased to 0.06 at the next level.
if you want to make people choose what to do with their specialists, then make them choose "one of the following" different level pre-req buildings. a new mechanic is not necesarry to do this. if you want people's production to suffer by taking people out of the economy to go fight, then do this by TAKING THEM OUT OF THE ECONOMY. ie, production goes down because it is dependent on population. if the numbers don't allow for this at the moment, then increase the population costs of troops (to 5 per guy or whatever) so that it does matter. this represents the fact that others are dependent on the guy you just took.
the value in being able to work together better on the battlefield should be in WORKING BETTER ON THE BATTLEFIELD, not in it's cost to society. inventing the phalanx or fire by rank did not allow us to field more guys. it just made them more effective.
personally though, i don't think making people choose between research and building units is fun, because it forces them to have lots of settlements to get a decent mix. if you really insist on speciation, getting people to rebuild their cities is not the way to do it.
in summary, stop trying to make abstract models for things and just have the game do them instead. stop trying to hit me and hit me.