Some collected responses below.
First one general point: I'm mostly talking about single player here, although some of it may apply to multiplayer also. At least in multiplayer everyone can use the same tactics, and there's fun to be had in doing it better than your competitors.
I completed disagree on the national wonders issue, I through they were much too powerful in Civ 4. BUt other than that I agree with a lot of this.
Remember I'm not saying whether each individual thing is good or bad, there's a lot of factors involved and ICS is just one. I quite enjoy national wonders in civ4 myself, but I agree they can dominate things and that might not be for everyone. The developers might have just decided we could use a change from a strong national wonder focus.
Reversing the point value of happiness buildings doesn't solve ICS, if we viewed ICS as a problem. It would simply reward a lot of big cities rather than an equal amount of small cities.
I defined a lot of big cities as something different from ICS. As Louis XXIV said it's harder to acheive. As a result it doesn't feel so degenerate. Opinions vary though, and not every alternative is considered better by everyone. Nor is every factor a difference between ICS and every alternative.
All buildings are single focus, and the only differentiating feature most of the time is increased cost & maintenance. There's no complexity or sense of dynamic to it.
It's a little off topic, but it made me think. Almost every single building in civ4 did at least two different things. I just had a look at the list, and there are very few exceptions. There's almost no buildings in civ5 which do two things other than wonders, and even then all you usually get is a measly 1 culture and 1 great person point per turn. Specialist slots are an exception still at least. It's a subtle way in which infrastructure is weaker, in that you get nothing for free. In civ4 you might want gold, and you could choose between market and grocer depending on whether health or happiness was more important. In civ5 if you want happiness you choose between circus and colliseum, both of which are identical except one gives more happiness for the same cost. The side effects gave a nice dynamic, and I can't see why they'd want to take it out, it just smacks of design laziness to be honest.
This is actually fine.
At the beginning of the game you need some extra happiness at times but economy is a lot weaker due to no banks/civics and so on than in later game.
I might be misreading this, but my point was not that the buildings are cheaper, because they need to be. My point was that they were not just cheaper but better as well.
The problem isn't with per-city bonus, it's with the flat per-city bonus.
Fair enough. I wasn't comparing to the % based alternative, because it's an alternative solution not a real comparison point. I've added the word "flat" though to make it clearer.
ICS might be the thing do to, but I don't see why players who enjoy having large empires should be penalised.
I agree. I think there's a difference between genuinely large empires and ICS though. Some changes might penalise large empires though and the developers need to keep that in mind. Other changes might help large empires and cost ICS - making making growth at high population faster and the later happiness buildings better for example.
One comment I would add about trade routes...
Now that trade routes are just 1+1.25*citysize, that 1

at the start of the equation is effectively a per city bonus now. Small infrastructureless cities (apart from a colloseum and perhaps a theatre) turn both a gold and happiness profit very easily. The 1.25

you get per population point is cheaper in smaller cities than it is in larger cities because it takes less food to grow to small populations.
Great point, I've added it to the first post as another factor.
And finally, thanks to all who've made positive comments
