Thanks for the feedback.
City assignment after collapse does take stability/settler maps into account, but also a number of other factors, including previous ownership and cultural influence. I agree that this is sometimes too powerful though, and we don't always want complete empires to emerge. Being too shy with this also has its downsides, however, because of the inertia rule, especially in earlier eras.
I agree that the other crises types are problematic, especially minor crises. More on that below.
I also like the idea of getting a minor benefit from hitting solid. Nothing as major as a golden age, but a temporary bonus depending on the category you're doing best in sounds good.
While I don't have the time to actually implement stuff at the moment, I've been thinking a lot about the stability system and what's working and what isn't. My current conclusion is that:
1) The previous ideal that stability calculations should only be made when a stability check occurs and therefore depend on the current snapshot of your civ's situation caused a lot of problems, because it meant the lack of long term data to base the stability on. The result was the unpredictability of the results in a number of categories. I have already moved away from this in the economy rating, and currently have ideas to do something similar for warfare and happiness (part of domestic).
2) There still are not enough ways to intentionally generate positive stability, or rather the ways that exist aren't properly balanced against causes of negative stability. There won't be a real paradigm shift here, I just have to put more effort into enabling this after I'm done with the changes from (1).
3) Expansion penalties for "normal" expansionism, i.e. the likes of historical empires, compared to domination victory expansionism. Punishing one of them appropriately either means not punishing normal expansionism at all, or punishing domination expansionism too much. And I want to do the former because controlling China as Japan should cause you trouble (for instance during an UHV game), even though China isn't nearly enough for a domination victory. I have a few ideas here, but currently I'm without a real solution.
4) The ubiquity of crises other than collapses and their disrupting effect on the game, and the relative rarity of partial collapses. I think the main job of the stability system is to check expansion. Partial collapses are a good feature in my opinion, but aren't applied often enough. Instead often other crises types get in the way. I will probably reduce the effect of these crisis types and make them occur less often, and only at lower stability levels. The intended result is that they're meant to give civs some negative side effects even if they aren't overexpanding, but to make partial collapses the main outcome of negative stability for those that are.
5) Positive effects from hitting solid have just made it into my list.
bonus from workshops and / or the 
bonus in the capital.
Quite a strong argument to switch to Guilds, eh?
Latest SVN mouse over Expansion box reveals negative modifiers for razed cities only. As Germany I own London, Moscow, Istanbul, Paris, London and get no negative modifiers whatsoever for owning foreign core. It seems that Civics now play the most important role in stability calculations 
