@Karavinka
Mmm something would have to be done about rebel spawning in hard to reach places I suppose. Since they're supposed to endanger the capital, it would make sense if they only spawn in a province that has a land connection to the capital. Would big rebellions (as in, a few 20k rebel stacks spawning in provinces near one another) be like goombas? Well, maybe if the goombas started breating fire... Sure, the country AI is smarter than the rebel AI but it still isn't very hard to outsmart it. Giving rebels large stacks would imo go some way to briding the difference between said rebels and national armies. As the OP mentioned, large-scale rebellions would probably be exploited the enemy AI declaring war on the empire.
As I have written in a previous post, I'd like the game to simulate the insurrections and the infighting that have almost always accompanied large multicultural empires historically. One more ambitious way of accomplishing it would be to add the rebel states from CK2 and Rome. I know from your previous posts that you(Karavinka) aren't overly interested in historical plausability. Therefore I understand if you feel that an added rebel mechanism is not the best way of increasing difficulty. I personally would on the other hand like any events/modifiers that increase late-game difficulty to feel historically correct.
I'm actually OK if the problems inherent in a pre-modern multicultural empire get represented in the late game. But to me, simply adding more rebel events sounds like adding 20 goombas per stage instead of making the level design and boss fight more interesting.
What I don't like about most arguments from history on the forum is that they often tend to be one-sided. You know one fact, and you want the game to represent that one fact, while ignoring often contradictory facts around. Like how some people claim "oh Russia is supposed to be overstronk don't whine it's historical"... yeah, I can totally buy that argument if Taungu also builds itself up to its historical extent in any 1444 scenario started game. "Oh Novgorod is meant to fail" yeah... why doesn't Vijayanagar not only collapse but also thrive in most games? What about Majapahit, should we have a trigger event to collapse it around 1500? Why isn't there a historical rivalry modifier between Ming and Manchu? They end up marrying each other and never fight since they don't really have anyone else (well, Korea) to marry. Does that make any sense? Of course it makes sense if people have their primary areas of interest, but to me, it just sounds fake. Well, I digress.
Sure, multicultural empires had a lot of problems, but before adding rebels, I think it's fair to add some measure that the players can act on. What I think is needed is a more elaborate culture mechanics itself: why can't we designate one or more cultures as privileged over others, for example, like Persian in many Indian Sultanates? (In my 1.1 Gujarat game, I had a Sultanate idea "Persian language at the court" even though Persian wasn't an accepted culture. And I had all existing Persian-cultured provinces. Hell?) Why can't we set one ethnic group against another, to keep their hatred toward each other instead of all revolting against the central government? Isn't it also realistic that, say, if Hessians and Rheinlaender provinces get conquered by France and they're engineered against each other in a rivalry, they both may want to prove themselves loyal to Paris to get a better treatment over their rival? Or, do we have immigrations and population movements of any sort in the game? The wave of nationalism started in Europe and did not really spread into the rest of the world in the EU4 time frame either, and a lot of people based their identity on the clan lineage or region rather than a "nationality." How do we account for these sorts of things? I'm not saying the revolts are ahistorical, there were revolts and resilient culture groups (look at the Aramaeans in the Middle East, for example) but that's just one side of the picture.
I'm completely OK if Ragusa conquers half of Europe in a chain of (apparently very lucky) events, as long as each step makes sense and is not likely to be repeated every game. But it's not that I think history or plausibility is not important - of course, I want the game to make sense, except that I acknowledge that the real world cannot be abstracted into a game "accurately" or even "reasonably well." Don't take it personally, and yes I think something needs to be done with the late game, but throwing more rebels sounds like a lazy solution to me that adds nothing but frustration.
EDIT: I guess I need to summarize that long rant to make myself clear. I want to propose a more elaborate mechanics with intuitive interface to deal with the cultures in your realm, like how one culture feels itself privileged over others while another feels itself oppressed, along with the measures that can be done to manage them: e.g. acknowledgement, subsidizing, patronizing, sowing distrust amongst minorities, repressing, exterminating, forcible migrations, etc. If the player screws up with the AE and gets into a large, historically impossible pan-world coalition, it's the player's fault. If the player screws up with the internal cultures and gets into a long dreadful rebel event chains, then it should also be the player's fault, but rebel events "just because" don't sound like an appealing design to me. Making culture management difficult is one thing, just throwing in rebels without giving you real options is something entirely different.