Microsoft is embedding Live services within the operating system. This isn't really too surprising, considering it was one of the stated principles behind the effort from the start. (As a developer, it is exciting though).
Microsoft appears to recognize that web applications are a complement, and not a substitute, for desktop applications. The goal is to deliver the richest experience available on a particular piece of hardware, and no matter how many times Steve Gillmor says it, that will never be in a browser.
In fact, the example I use most often to illustrate the point above comes from Microsoft's Exchange Server. If I'm on a machine that has Outlook, I'm invariably going to choose that over Outlook Web Access - but I appreciate being able to use the rich OWA interface when I am on a machine with just a web browser. And if all I have is my mobile device, Outlook Mobile Access offers a simplified interface. The important thing is that, in each case, I'm accessing the same data.
I think the next generation Office products will work the same way. Not just that, but if I had to bet on either Google or Microsoft winning this, my money's on Microsoft. (Disclosure: I own shares of both, and I don't think either is going to "lose"). Why? They already have the great, rich client application and they have the technology (Sharepoint, Groove) to offer the hosted access. The only thing left is the web application and that's trivial to build.
I also think more of these web applications will be self-hosted in the mold of Exchange Server. There will still be hosted versions available, but more individual users and small organizations are reluctant to trust third parties with their sensitive data.
In related news, Martin Taylor, the executive leading the Live marketing strategy, has abruptly left Microsoft. There are no details on why, but the language suggests that, if it was not mutual, he was pushed out (e.g., "we have made the difficult decision to part ways").
Windows Live was a great technical shift, but I think it was positioned and marketed poorly. (If it was marketed well, Steve might actually understand why it makes sense and that Microsoft is not so pin-headed after all.) Instead of trying to position against Google and Yahoo, Microsoft should have emphasized the integration with Windows and all the benefits that come along with that.
But then, we all know marketing is not necessarily Microsoft's strong suit.