Great! Quickly, everyone immediately stop all software related development that doesn't have a visual or performance effect that is directly obvious to the end user. You may only develop on XP from now on, you may most certainly not create new frameworks that will help speed development in the future, it doesn't help your customers and isn't worth it.
Don't make the mistake of thinking that Vista was built soley for the purpose of advancing computer technology/software engineering. It does a bit, but it comes at a price. It certainly advances DRM technology, for example. Is that a good thing? MS are a corporation too - they're out to make money, and frankly, are only interested in developing new frameworks if there's a few dollars in it for them.
It's an ok OS, but I have to point out that at least in my experiance, it has had poor driver support (slowly getting better), a few serious bugs on release and incompatibilities with some big apps. Now I aware that a lot of the fault isn't MS's, but at the same time, they did market Vista very heavily, when in fact a lot of users didn't really need it.
And anyone who says it's faster than XP in DX9 hasn't done any benchmarking. In that regard, it's still more of a downgrade than an upgrade. Hopefully that will change, but it's happening very slowly.
Having bought Vista on release, the fact that I still use my XP install for most games and some important apps speaks loud. Don't get me wrong, I'd love to finally get rid of XP and have just one OS, but Vista simply doesn't fill in all the gaps atm. That's not a good thing.
Back to the subject at hand - there's a lot of debate out there about weather DX10 could have been implemented in XP. I think if it was, developers would be more interested in it, and that alone would helpt to advance the framework much faster.