I've written on this general topic before, but this latest news puts things into a slightly clearer perspective for me. WinFS, Avalon, .NET, etc. are all now either dead or deeply decoupled from Microsoft's next operating system release. Their new major OS, Vista, is not for consumers, its maybe not even REALLY for developers (as I'd previously argued) - its for businesses; it'll help corporations with their TCO with regard to Windows.
I'll argue that's really what "security" is code word for, in the context of Vista: "cheaper to support". If it helps anything else (like users :P), that's a side effect - a bonus, really.
And that might not be a bad strategy for Microsoft. The single biggest threat to their platform effect is OSS - Open Source Software , along with the platform of the Web generally - which removes distribution as a significant barrier to entry, and comes replete with its own eco-system of tools and maintenance - the valuable Intellectual Property chain of TCO.
And the math there, as a very smart friend at Microsoft recently pointed out to me, is really simple:
MS License $ + MS TCO $ must be less than OSS TCO $
There's little doubt that Vista will be a giant cash cow for Microsoft - its been 6 years since they released a substantive upgrade to a core business product, and coupled with the 3 years since a new version of MS Office - well, I think people have forgotten/underestimated how much cash will be cycling next year in Microsoft's wake.
Unfortunately for all of us, as consumers, though, it feels like Microsoft has lost their way a little: they're now in the business of being in business. All things in business cycles degenerate to spirals, positive or negative - and reaching the Ouroboros phase of self-referential consumption is a dangerous place to be.
As we can attest to at AOL, getting snared by your own success is the most insidious trap of all.