Actually, there is reason to think Win8 adoption will be slower - many people were still on XP (a 10 year old OS!) and holding out on upgrading to Vista because it was so bad (perception wise; no comment on any objective measure of quality). So when Win7 came out and was not bad (perception wise), all those who held out were happy to update. It's possible the Win8 updates won't be as popular because Win7 (like XP before it) is good enough that there is no compelling reason to upgrade.
But even assuming it does do 300 million in 2 years - poll your favorite developers; what is the time frame since win7 was out that they are willing to do a "win 7 only" release? (no vista support, no xp support)? My bet would be that maybe some would NOW, but no one would have a year ago.
.NET has been generally available for 10 years now, .NET a part of XP/Vista/Win7 for over 5 - and the vast majority of generally available desktop software is STILL not using it. Why would you believe a new Win8 runtime would be different?
Why haven't Windows 7-only applications caught on? A major reason is consumer expectations coupled with only a marginal effort required for backwards compatibility for a big increase in market. .NET was a technical back-end change that isn't user-facing, so it competes with legacy code-bases and existing experience.
If Microsoft's new framework fails, it'll be for different reasons.
And that's reason enough for Win8 applications not to catch on, unless microsoft manages to create a substantial "other" market in the form of phones or Xboxen. Whether or not they succeed, we will have to wait and see.
Apple proved that you don't need the largest installed user base to become a success. You just need a user base that's large enough to be sustainable, coupled with decent developer tools and, preferably, a sales channel.
That also presumes there's no backporting, which is another possibility.