Paul Thurott explains why I'm not worried, using the kind of argument I've brought up before:

Microsoft has made some mind-numbing mistakes. It (illegally, as it turns out) artificially bundled its immature Internet Explorer (IE) Web browser so deeply into Windows in order to harm Netscape that it's still paying the price for the decision--a full decade later--in the form of regular critical security flaws that have taken away time from developers that might have otherwise been spent innovating new features. The company itself has turned into that thing it most hated (read: IBM), an endlessly complex hierarchy of semi-autonomous middle managers and vice presidents of various levels and titles, many of whom can't seem to make even the smallest of decisions. The company is too big and too slow to ship updates to its biggest products. It's collapsing under its own weight.

To make a long story short, Microsoft has become Gulliver, tied down by a million tiny threads. Unlike Gulliver though, their personal Lilliputians all work in Redmond, endlessly creating more thread. Which is not to say that Microsoft will disappear; you'll notice that IBM isn't gone. It does mean that MS faces some very rough seas, and their level of industry influence is on the wane - and I very much doubt that it will cycle back up anytime soon - if ever.

Hat tip Keith Ray.