At one point their system was working. But then they went and broke it again.
The primary factor in it's workings was... economy. Maintenance namely. Once upon a time building those happiness improvements REALLY costed you. It was actually in favour of tall empires a bit. But then they had to go and rebalance the numbers. THEN building those now MANDATORY happiness improvements stoped being such a notable detriment, and it all inevitably went down the drain.
The biggest part of their "tall vs wide" balance was the combined effect of "per city" improvements and "per empire" factors (such as luxuries).
It's a working concept and with the numbers set right it works right. But for some reason that eludes me to this day (ok, let's be honest, I suspect that it's something along the lines of BuckGodot's comment quoted earlier in the post), they went and ruined their own math...
Not really, no. ICS was still better even in post-patch, pre-g&k vanilla; you just had to keep an eye on population. It was only really killed in Civ 5 when they implemented the research penalty for new cities. From release right up until BNW, 'tall' was always weaker,
While at the post-patch, pre-G&K stage they succeeded in making happiness act as a general 'population limit' which could be spread through however many cities, Global Happiness never managed to deal with the other advantages to 'wide' empires which 'Tall' ones couldn't match. Wide empires grow faster, and so can fill up any new happiness bonus faster. Wide empires have more build queues, allowing them to do multiple things at once. Wide empires have a larger number of tiles per population point, making them more flexible. A larger % of a wide empire's population is covered by the free food from the city tile itself, so you're paying less support for the same amount of population. Population in multiple cities is inherently more valuable than population concentrated in one city, and was right up until they put in the research penalty.
It's pretty simple, really. You can use a mechanic to challenge expansion OR development, but you can't use the same mechanic to challenge both, because the strengths of the two approaches are different. One will be better at dealing with any given mechanic than the other, more or less automatically. Trying to balance one mechanic for both is therefore incredibly difficult, and also pretty much futile. And that's why, after half a million patches and 2 expansions, Firaxis gave up trying to balance it around one and added the research nerf. They finally just gave up on one-meter-to-rule-them-all and added another mechanic to limit expansion.
Which brings us to GC3. Presently, 'Tall' empires are behind on more or less everything. They have worse approval until LEP gets ridiculous, because they need larger populations and the approval equation involves DIVIDING planetary approval by population. If you colonize 2 planets and don't build any farms, then your total population will max at around 12 and approval will sit at around 50% without any buffs. On 1 world, it's down to 25%.
If you have 1 world, to reach 12 population you must:
a) waste a ton of production and multiple tiles on farms
b ) still have to build the same number of happiness buildings as the other guy
c) take twice as long to reach the population cap, since he's growing at the same rate on two planets - or spend yet more production on a hospital
d) have half as many tiles to work with
e) have less free production from the colony capitals
f) have fewer build queues
g) Need to build factories to catch up with the additional production from the colony capital
h) have to pay maintenance to get the same resulting food, production and growth from the same population.
You're basically screwed on every limiter. Maintenance doesn't effect width at all but it will punish you for every inch of height. Approval hammers your large populations much more aggressively than the additional penalties from LEP effect wide empires for the same population value. You're choosing to live without the free extra growth and production; you should be rewarded for that sacrifice rather than punished even harder for failing to expand.
On top of which, let's just look at the penalties that the big empire is suffering, shall we? Assume he's got enough LEP to wipe out his colony base approval altogether (presently, that's 15 planets) and has somehow reached that total without gaining a single approval bonus from anything:
a)He'll suffer -25% growth. This does pretty much nothing. Since growth is generally around 0.2 (very rarely higher), 25% is a rounding error. Besides, he's gaining 1.5 pop per turn empire-wide, so he'll still produce a full planet's worth of new production every 4 turns. The %-based growth bonuses really achieve almost nothing when growth sits determinedly in the low 0. range.
b)He'll suffer -25% production. This might seem painful, but really isn't, since again he's receiving far more than that in free production from his many colony capitals.
c)He'll get 25% less influence, but again he's receiving more than that free from all the colony capitals - and it's coming from a dense overlapping lattice of planets.
d) And he'll have 75% less resistance, though I've never actually seen an invasion fail either way. Also, if the enemy can actually get a transport to an undefended planet when you have dozens more planets than him, you're actually just bad at this game.
That's it. That's his burden for 0% approval (and even then, only if he doesn't have Patriotic otherwise, he's on 50% and has no penalties at all). And the Tall empire is probably suffering that penalty anyway, since he needs to build up huge planetary populations if he wants to remain remotely competitive and then has to use lots of tiles for all the approval bonuses. Plus, even if he's stuck to just his homeworld, he's still got some LEP himself, too, so his population aren't that happy either. Oh, and just to add insult to injury, given 50 turns the wide guy can STILL ramp up every single one of his worlds quickly and simultaneously to be a match to the best class-equiv world in the Tall empire.
Basically, happiness should limit tall empires and prevent them from concentrating too much power on one build queue. Maintenance should limit wide empires and prevent them from getting lots and lots of build queues - not by imposing an artificial limit, but by making more and more of the planets hosting those build queues focus on cash production instead of industry.
This is easier to balance, since you can now alter one limit without affecting the other; it scales more effectively, since planet sizes aren't changed by the map size and maintenance can be made to grow per-colony; and it just plain makes more sense (Empires have bureaucracies which cost ever-increasing amounts of money to maintain, but aren't necessarily consumed by sadness - it's particularly grating in 'good' empires where you've been lovely to your population all game).