Not to be a pedant, but a nova doesn't "die out" within the distance of 1AU.
Novae and supernovae aren't just bad, they're superomgwtfbad. They destroy their own entire solar system, and usually the solar systems nearby, and some emit enough radiation to sterilize pockets of the galaxy around them. A supernova going off near our sun would be as bad as say, activating the halo weapon system. It'd annihilate mankind as surely as the sun rises.
The kernel's driver system is largely the same as Vista's to avoid that enormous hurdle Vista posed to hardware companies in continuing their product support.
This is why initial reaction to 7 is so overwhelmingly positive: the majority of the problems with Vista in the first place were third party drivers.
7 does have a more modular kernel, better memory management, better lock management, a cleaner and more thoroughly tweaked UI, and other benefits. I don't want to dismiss the work the MSFT engineers did on the kernel, as they've done -a lot- to improve it and those improvements will only really show in certain performance benchmarks. For the most part, you and I won't notice it.
I'm going to be installing it on my desktop on an empty partition tomorrow evening, and I'll let you guys know how everything runs. Initial reports lead me to believe that there are essentially no compatibility problems with Vista, and all the features in it just work, without any problem.
Bottom line is, unless tomorrow's install experience contradicts the widespread public opinion and benchmarking, Windows 7 is what Vista should have been, but could not have been, due to that driver change. No matter what, Microsoft was going to have a problem there, whether they introduced it in Vista or 7 or 8 or whatever.