Google's Android platform is no doubt an impressive and rapidly maturing mobile system strategy. But I think Google is falling into a trap here.
Google grew up as a server-side web company. Anyone who's built software for the web knows that one of the great benefits is being able to control the whole experience, and update that experience whenever you want, since everyone accesses your world through a (relatively) consistent web browser engine.
But with Android, they are facing a whole different beast. While being praised (especially by the anti-iPhone crowd) for being “open” and for licensing their OS to all mobile providers (not just phones, but also tablets, netbooks, and embedded devices), they are fracturing their software far and wide onto devices they don't control.
Microsoft fell into the same trap. (Windows runs on just about every piece of commodity PC hardware out there). It's an understandable siren-song: the promise that you can just make software and license it, and let other companies worry about the challenges of hardware design, manufacturing, consumer marketing, etc. Software is a high-margin business, its pull is strong.
But Microsoft is suffering from that decision today. They have the largest user base of any software company in the world, but they've completely lost control of it. Tens of millions of users are running hardware and software revisions from more than a decade ago (and hundreds of millions still on Windows XP), and all that legacy hardware and software creates enormous overhead for Microsoft while trying to innovate Windows. It creates a mess of driver compatibility (that almost single-handedly ruined the entire Vista proposition). And frankly, it's just confusing for consumers.
Google is going down the same path, with a bravado and apathy that only comes from a giant, wildly successful, practically brand new company.
Furthermore, Google is letting developers release apps into various app stores with little to no consideration for the end-user experience. Developers don't have the resources to test their applications on the dozens of different Android phones, so those apps will just break on a lot of phones. Consumers are left to wonder whether an app they download is even compatible with their device.
All of this is good in the honeymoon phase. Microsoft has proved it can be a pretty lucrative strategy. And I think Google will shortly dominate the handset market, passing Apple's iPhone juggernaut comfortably by the raw numbers. But I suspect that in 5 years, consumers are going to find Android to be a confusing mess of a platform, oodles of incompatibilities, carriers who have customized (read: broken) the experience on their individual phones, and developers will hate being forced to develop for a platform that's become too ubiquitous to ignore.