The fatal flaw with the current business model.

While this may (or may not) be true, you have to realize your talking about the future here. It is unlikely that they would fail so suddenly, there would most likely be signs. Of course, when those signs do appear, it would probably only worsen the problem as everyone jumps ship.

The problem is that Nexon is pretty terrible at recognizing such signs, as they have repeatedly demonstrated.
Also I wouldn't say that the failure would be sudden, considering it's pretty much already happening, based on peak concurrent user data and the quarterly reports.