I'd agree with some of the analysis of the strengths of open source development, particularly that liberal (BSD/Apache) style licenced projects are winning out over GPL ones.
But a large reason for that is precisely because the likes of IBM, Apple, and Google, are investing heavily in their success. I don't see how you can leap from that to :
"Ultimately, all single-vendor innovation will be commoditized through a community-owned project."
Because that is an utterly unproven theory. Even a project like Firefox is 'vendor' driven (although the vendor in that case is an independent foundation). WebKit is dominated by Google and Apple. Ditto the HTML5 standards process (WHATWG vs the stalled W3C community). There's similar issues with the JCP.
I'm not saying community ownership is impossible or always doomed to failure, but I don't see any evidence that it is a natural next stage in the evolution of software.
Equally, I'd disagree with the idea that 'openess' is more attractive to developers - otherwise, how do we explain the hundreds of thousands of developers working on top of proprietary systems from Microsoft, Apple, Oracle, etc. Except that of course, they can't be 'real developers' if they don't agree with the cause.
For what it's worth, I do believe Android will become the world's dominant device OS - but only because of Google are driving it. If Google pulled out, the mobile vendors would likely make the same mess they did with Symbian.
(Which is another good point - at the point the iPhone arrived, we'd had a good decade of Symbian and J2ME - both mobile vendor driven community efforts. By the theory, they should have been well ahead of a proprietary latecomer)