quick question: what is the 32-pop buffer? (does it mean not to exceed 32 pop?)
Just a hypothetical.
For tradition, you will always be able to obtain more happiness, so the focus is grow grow grow.
For liberty, your happiness is capped somewhere, depending on your religion, social policies, availability of zoos/stadiums, city states, etc. This "cap" gets majorly lifted when ideologies come into play, as you start being able to select 3-4 happiness policies in a row if you so choose, at not so high opportunity costs (because besides 1-2 policies, the happiness policies, especially tier 1 ones, are some of the highest value social policies in the game). But, until then, you have to distribute your population efficiently, which means you don't necessarily want your 7th city to grow to 10 pop (yet), even if it could. In fact, because all happiness can be traded for gold, if you are pursuing a culture strategy, you may not want to grow those cities at all (stay at 4-5 pop) because the reason you added more cities was not for science, but for archeologists, or space to put great works, or maybe you have a UI that translates into tourism, or you're just banking on getting more dig sites in your territory.
Anyway, say we're at zoos, each city has +4 possible happiness, we have +2 for religion with temples, and +1 for a road. That means each city is happiness neutral at 4 pop. You then have a global happiness "bank" of luxuries and city states and what not, and that is the max additional pop (+5% with Liberty) in your empire for you to distribute. You'll want enough pop in your culture city to support the specialist slots, if you are producing things, you'll want enough pop in that city to be able to work mines (or trade posts with a gold city), but most importantly, you'll want population in the NC city. If you have a total of 40 extra pop and 8 cities, the worst way to divide them is 5 extra pop per city. Your cities clearly should be prioritized differently.
The actual # of the happiness buffer/bank changes depending on the game, but you should have a rough idea of where it is, so you can do some empire population planning (something completely unnecessary in tradition, where you're only limited by the amount of food you can pump into a city).
Because of the price you can sell luxuries at, each pop costs you a static 2 gold per turn, so long as you have any luxuries you haven't sold/traded yet. In early game, that's hefty. In mid game, that's the same as a military unit, or worker. In late game, it's not even half a worker. Justice is making the point that at a certain point, pop is no longer worth the cost of 2 gold per turn and the hammer opportunity cost to get the pop in the first place, because all of it's value (besides science) is being put into food (which only gives more pop) and hammers (which is being used to build happiness buildings / food buildings / hammer buildings). For a science game, I don't see it that way, since you would pay almost any amount of gold for more science. So, all I'm worried about is the mid-game happiness bottleneck and thus proper population distribution (this also means it's not necessary to run all those food routes, just 1-2, so you end up with more gold for mid-game).
