MS spent as much effort on the UI as they did on the actual product. This is very different than FOSS.
This is a key problem with FOSS right here. The developers don't see the UI as being a core part of the actual product. That's fine, I suppose, if you don't want users to interact with it, but if user interfaces remain an afterthought rather than a focus, then most users who aren't themselves developers (or among the fringe that chooses FOSS on principle) will rightfully say that the FOSS alternative is a vastly inferior product. For the user, the UI is perhaps the most important aspect.
Microsoft's strength has long been in user interfaces. They have the resources to do vast amounts of market research and user testing to develop interfaces that most people can interact with and actually get desired results. FOSS products are generally produced by developers who are most interested in how to code the underlying functionality, and a GUI is just slapped on as an afterthought. Products that are successful in the market are produced by people who view the UI as being of utmost importance, not an add-on to "the actual product."
Even in proprietary software, look at Windows Vista and Windows 7 as examples of how this plays out. With Vista, Microsoft concentrated on rewriting the entire OS in modern programming languages, changing the underlying structure of how the OS utilizes memory, etc. UI changes were made, but they weren't the main driver behind the product. It flopped, due primarily to bad press resulting from some major marketing gaffes (allowing Dell, et al to claim PCs with insufficient hardware were compatible) and drivers not being ready yet when Beta/RC versions were released. The product, as it exists now, really isn't as bad as its reputation makes it out to be. But since the user experience wasn't the focus, some significant flaws there have tarnished it. The perception of the problems with the user experience in Vista are in part responsible for the market share Apple has been able to gain in recent years. It's certainly the path of attack they've taken in their advertising.
Windows 7, on the other hand, does improve on the performance of Vista and incorporates the bug fixes from its service packs, etc. But its main selling point (other than being "not Vista") is a handful of simple, but significant, UI changes. The underlying core components are essentially the same as in Vista, with the improvements from Server 2008 thrown in. But the focus was on the user experience this time around, rather than technical details, and the results are apparent. Nearly everyone who's tried the release candidate is giving it rave reviews, and the currently prevailing opinion seems to be that it's likely to be a hit in the market. The reason for this is that Microsoft focused more strongly on the experience of the user, both by adding some slick new features and by eliminating the major annoyances (obnoxious UAC, slow boot times, etc.). Some of those issues are related to underlying technology - user experience is not solely a UI issue - but unless the product is made with an eye towards how the users will interact with it ... well, they just won't.
Whenever the user experience is thought of as secondary by the developers, users will notice that they're being condescended to. For some applications, e.g. Apache, which are primarily used by server administrators, it's not always a deal breaker, but for anything "on the desktop" it's critical.