Tim Lee makes a good point at Freedom to Tinker - that open systems always seem to be losing until they've won - in part because the narrow interests that favour closed alternatives affect the reporting of the battle more than they affect the battle itself. ("The grassroots users of open platforms are far less likely to put out press releases or buy time for television ads.") But the open systems win for the reasons I talked about the other day in the context of a "child-friendly internet" - the closed systems seem better fitted to customer demand, but they don't adapt the way the open systems do.
Lee goes on to say that open systems will always win. I'm not so sure about that. Open systems will continue to win for as long as adaptation and innovation are crucial. When (if) the service required by users becomes stable, it seems probable that open platforms - with their complexity and vulnerability - will be supplanted by black-box single-purpose "appliances" that Just Work.
How likely is that to happen in the realm of information networks? I don't know. Possibly we'll always want more information. There may be subsets of networked information that can be hived off onto closed platforms, but against that, there's always likely to be a value in combining information, either on your own systems or upstream. If some of the information you want is on open systems, the rest will need to be able to interact with it.