When KDE 4.0 was released, there was a lot of criticism being thrown its way, but it's important to make a distinction between two types: the criticism that KDE 4.0 shouldn't have have been released as 4.0, but as a developer preview or something - and the criticism that KDE 4.0 was different than KDE 3.x, and did things in a different way. Windows 7 didn't have the released-too-early problem (yet), but did face a storm of criticism from people like Paul Thurrot for being too different from what came before it.
In a sense, with today's release of KDE 4.2, both KDE and Windows 7 are in a very similar position. Both KDE 4.2 and Windows 7 build upon frameworks and technologies that have been available for a while, exposing them to users in a useful way. For instance, Windows 7's HomeGroup doesn't introduce any new technologies; it builds upon low-level features and frameworks already present in Windows Vista. It just ties those low-level features together and makes them transparent to the user. The Windows 7 taskbar is a similar case; it also ties features and abilities of previous Windows releases together to make the taskbar a better fit for the Windows application/document paradigm.
The same more or less applies to KDE 4.2, only to a much greater degree, as the changes between KDE 3 and KDE 4 are much more encompassing and drastic than those between Vista and Windows 7. KDE 4.2 builds upon the frameworks and low-level features introduced in previous KDE 4.x releases,and exposes them to users in a useful way. This process is on-going, and future KDE 4.x releases will only expose more of the fancy new frameworks to users.
Both KDE and Windows should be commended for trying to change some of the established UI conventions, trying out new ideas, and implementing new approaches to user interaction. KDE's ideas are much more drastic than Windows', but the essence remains the same.
Like I said, I've seen many people criticise the changes implemented by both Windows 7 and KDE 4, and all of these criticisms seem to come from people who at least call themselves technically inclined and computer enthusiasts. I realised ages ago that computer users most averse to change are not normal, every-day users - but nerds, geeks, computer enthusiasts. Normal users aren't as averse to change as many seem to imply.
Hundreds of thousands of people switch from Windows to Mac OS X every year. I've seen many of my friends switch to Mac OS X without so much as a hitch. A few pointers, maybe, but that's it. If you've ever seen someone use Microsoft's ribbon interface for the first time, you'll see that it doesn't take them very long to adapt.
My point is that us tech folk always criticise the normal users for being averse to change, inflexible, and, well, plain stupid, barely able to find the "on" button - but if my years and years of OSNews reading, commenting, and editing have taught me one thing, it's that the more technically inclined people are, the more averse to change they are.
No better is this illustrated than by Steven J. Vaughan-Nichols latest piece. He criticises both KDE 4 and Windows 7, but all of his criticism basically comes down to "I have to do things slightly differently! Help! I'm lost!" - exactly, the same thing we are so eager to blame normal users of doing. And I might have singled out SJVN, but he's not the only one of his kind being averse to change. Just browse through the OSNews comments' section to any article related to KDE 4, Windows 7, or Office 2007, and you'll see one after another.
I find it slightly amusing that the people who are the most stern advocates of normal users moving away from Windows, trying out alternatives, are the same people who are usually lost whenever they themselves have to change their way of doing things. Do as I say, but not as I do.
KDE 4.2's release is imminent, and while it certainly isn't perfect by any stretch of the imagination, the KDE team deserves praise for at least having the guts to go against the stream and come up with something truly new. Instead of condemning them for it, why don't you give them an honest chance for a change, and not curl up in a fetal position when you discover they moved your pet feature three pixels.