There are two arguments against allowing complete collapse during peace-time. First, the gameplay argument, which has already been rightly stated: stability management is opaque and complicated enough that it must be learned in practice, which shouldn't be punished with "out-of-the-blue" defeats; there's a reason why modern roguelikes don't use 20-hour campaigns. Permadeath is fun, but only if, soon afterwards, you get the chance to show that you've learnt something.
Second, and perhaps more convincingly, there is the historical argument. While I'm sure that you can find some examples of civilizations at least possibly collapsing outside of war with a foreign power (such as the puzzling fate of the Minoans, the Classical Maya, arguably some dynastic transitions of power in e.g. China), I'd wager that this total collapse is extremely rare compared to a "collapse to core area":
* While it is a stretch to consider Alexander's empire a continuation of Macedon (arguably it was a successor state of Persia more than anything, simply with a new predominant culture, but this too I wouldn't quite subscribe to), either the Antigonid remnant (which afaik kept to the ambition of restoring the undivided empire), Macedon "proper" (the most likely choice for DoC, given that the "Dachzivilisation" is Greece), or even the Seleucids (because they held Babylon, which may have been Alexander's intended capital) could be seen as the "collapsed core".
* Rome during the Crisis of the Third Century suffered from overextension, with empires splitting off that could be considered "civ-sized" by DoC standards, but the core remained
* European colonial empires generally collapsed-to-core (Spain, France, Britain). Even the loss of e.g. India never plausibly endangered the continuation of some manner of British state on the Isles. Likewise, even the independence of Scotland by referendum (not implausible in the future, I'd say) would most likely not cause England to splinter into city-states with Manchester under different sovereignty from London (yeah, I know about those "London could choose to secede from the UK" sentiments)
* The Mongols also collapsed-to-core twice, first with the secession of the Golden Horde (or rather its predecessors) / Ilkhanate ~1260, then to the "historic core" with the loss of China 100 years later. It didn't cease to exist as a sovereign state (some subservience to Oirats aside) until the Jurchen conquest.
Simply put, complete collapse due to overextension is implausible because no matter how small your "predominant" culture is in relation to the whole empire, it will always have its core area in which it will have an interest in a national state to continue business as usual, if perhaps with a change in political party or system. Even gameplay aside, collapse-to-core should be favoured in peacetime / civil war on historical grounds.
Another argument is that polities that were distinct in sovereignty can be represented by the same civ in DoC anyway (Greek or Italian city-states, for instance) if their culture is similar enough (and there was certainly a lot of war between Italian city-states, e.g. the Guelph/Ghibelline stuff incl. the War of the Bucket "lol"), so using total collapse to represent a loss of central authority doesn't quite jive
haven't played DoC in years, though (computer can't cope anymore), but think that this has always been one of the pre-eminent issues with RFC (!)