'Relatively' can be 'yearly' with a sufficient population group. So long as the rest repopulate and resettle fast enough, it's a survivable pandemic.
False. If the society is broken up into small tribes or hamlets or the like, and there is no or limited contact, then what I said applies. Not only that but you need to review your history. When our own earth was in the stone age, total population was considerably less than a million. Total.
*Citation needed.
It's called math. Say you start with a total population of X (doesn't really matter). Let's say that half of the tribes are wiped out by abominations. Assuming there is no resupply (more on this in a moment), you get a classic half-life decay curve. After five generations, you'd have 1/32th of the starting population.
Now, let's assume (since this is normal for stone age type societies), that interaction between tribes is rare to almost unknown, and what little contact there is might be a handful of people...say no more than 1/20th of the tribal size. Now for a typical stone age tribe, you'd have about half the population of the tribe be of breeding age, and not all of them (admittedly most but not all) will be paired up. Let's also assume that tribes don't split apart until a critical population mass is reached (say about 150% of the normal tribal size).
The 'p' value (the NET number of surviving children per breeding couple) has to be 2 just to replace the breeders. It has to be a bit more than 2 to replace everyone in the tribe. If the p value is 3 then you get a surplus of 1 extra person per breeding couple....and 3 is a pretty typical net value. Given that half the tribe are breeders and given that you increase this number by half per generation (or 1/4th the entire tribe), you'd need two generations for N tribes to double (just to double the number of tribes...not double the population).
So we have half the tribes being lost per generation, but (pro-rating it), we only get 1/4th of those tribes back by breeding replacement.
Verdict: Extinction
Or, and this is simpler, we don't have to have extinction-rates of abominations to have socially unacceptable rates of abominations. Problem resolved.
It would hardly be a novel concept. Societies frequently refuse to tolerate things they could mathematically endure.
Even if the rate were low enough not to insure extinction, it would leave lasting and terrible scars on the suriving societies. Only ONE society shows such scars: The Qunari
Even a modest abomination rate (if it destroyed tribes) would be enough to prevent a magical society from ever evolving...yet we have Arlathan and Tevinter.....Hmmmm....
The Dwarves didn't have mages. The ancient elves all were mages. Tevinter was ruled by mages. The unorganized tribals had no means to systemically isolate mages, when the mages weren't already rulers or privileged classes.
Which means the Elves should have been a smoking ruin long ago.
Nobody required mages to be isolated from mundanes because before the Chantry, no one but mages had the power to do so, and they were the ones least interested in social division. They ruled society. That abominations rates didn't wipe them out doesn't really mean much of anything except that abomination rates didn't wipe them out.
Mages ruling society absolutely wasn't at all the universal rule. Certainly Andraste herself though there was nothing untoward with mages living alongside mundanes while cautioning that magic should serve men not rule them.