What I find most disturbing is the impossibility of getting out of the trap, once you were set. As you've described, the unhappiness caused some of your cities to secede, but instead of relieving your unhappy people, this just made them more unhappy, cause the big problem was in your capital.
When this happens for overexpanding, losing one or two cities actually alleviates the unhappiness.
The declared intention on luxury happiness was to punish both overexpansion and overgrowing. In principle, it's not that bad to have a mechanic telling the player that the followed strategy is not allowed, before it's too late. I mean, we don't really want that conquering city by city is all bell and whistles, there has to be some price or either the game outcomes would be too variable and it would make almost impossible to design any balance for victory conditions. The cost is there: too many cities and tech/policy/tourism costs increase too much. But before you get to a crazy expansion that criples your civ to a point of no return, happiness is there telling you to not take further cities.
Now we do the same with overgrowing. Growing too much may have an opportunity cost (that's why I think ElliotS does not see a problem with letting people grow to their hearts content), but it's also a winning strategy without the happiness limit (free yields everywhere, all specialists being worked, and all your territory being worked) and it's strong in the player's hands. VP happiness has always been like this, limiting tall playing in some way. I think it's ok that happiness hurts the player if growing too fast, but there has to be also ways to come back to positive, or at least giving some time for reaction. (Not a thing that can be said about surprise AI attacks). In your case, the best way to get out fast of the deep unhappiness would have been to lose some citizens in your capital. Losing satellite cities only seemed to make it worse.
Maybe better to take a look at the revolt mechanic, so that when the biggest unhappiness comes from the capital, it loses a percentage of its population instead of having some cities to secede. Maybe unhappiness could affect food production first. This way, even if the player makes mistakes, it's just some missed productive turns. Actually, I think this could be the best solution.
Then we could debate about how big should we let India have her cities.
