I am as to say a KDE “die hard” fan, but I am not a programmer, not an IT professional, neither did my qualifications include programming (leave aside some small statistical stuff). Still, since I am using KDE3.5 still, my view might me interesting as well.
So I do use still 3.5, is this because I think 4.x is no good? No. Is it because I want to visit my friend Lazarus and therefore I enjoy riding dead horse? Neither.
The current reason is that, for the time being, the 3.5.10 implementation of KDE fits in comparison best my current needs of functionality and stability. This wasn’t always so, when I think back, when an update from 3.1 to 3.2 was available, one was more than willing to update. But at the time there was no viable alternative for somebody who wanted to use KDE but wasn’t satisfied with the change. Now, THIS point changed and funnily enough a lot of people who where convinced that fear of change was the problem here, nearly “religiously fanatic” tried to push users who did not want to change …to make them change (remember "KDE3 repositories are not maintained anymore and the consequent “apocalipto” scenario for users who still insisted?). But THIS was not the point.
I tried 4.4 and the 11.3 on several recent installs I did to friends and “casual Linux victims”.
On my own PC I noticed that mainly (with some drawbacks in functionality) a non very advanced user will find now an acceptable system. ESPECIALLY if he did not experience the final status of KDE3.
Now the question if I well understand is: quo vadis KDE? The fact is that the perception of the MEANS of the DE is IMO totally different, when it comes to aspects as usability, completeness, ergonomics.
The KDE developer appear (in my eyes, please, this is a totally personal perception, and should be accepted as such), as programming a technical showcase. Currently they are interested in bringing the state of art of desktop programming to a technical level where they think to be able to achieve excellence. Therefore the change to Akonadi, the semantic desktop, the try to centralize the user data and part of the application functions into the DE to make it easier to create entirely new interfaces and to avoid duplication (currently a big problems of nearly all DEs). Now in all this the aspects that for us user are of uttermost importance, like stability, usability and completeness are absolutely by-products. Because what matters first is to get the fundamental architectural showcase running. The will surely come forward and may also faster and more interesting then ever before, but the way is still long I guess. (As it was long with KDE3 by the way).
All this raises in my eyes the impression that the problem is that the KDE developers have been “overrun” by the user inflow and the success of KDE3. While the KDE2 userbase were nearly all hardcore programmers and IT enthusiasts and barely normal professionals (leave aside even “noop users”) in the search for a stable and functional desktop, the current KDE3 and 4 user was (is) much more often that they expected a user that is not willing to beta test all the time. Neither for money nor for free. S(he) went away from a commercial environment disgusted by instability and was satisfied, expecting that the KDE3 stability status was something “permanent” and that regressions in functions where improbable, if not illogic.
And now these users are faced with the long startup curve of 4.x. Linux has success, and the user base changes with success.
I personally expect to change to 4.6 in march. I will have an outdated system by the time. But if I see that it will not be functional than I may learn how to compile a new kernel, leave aside that I will have to maintain some critical app. I hope I will not have to do it.
When I look at 11.3 I do not know if to blame OpenSUSE or KDE for the things I find unacceptable or incomprehensible:
- blue-tooth still not working (that makes 11.1, 11.2, 11.3 - and they talk about cloud strategy (shake head)
- Networkmanager and VPN support are (for what I have tried) still not at hight
- DVD/CD support of some applications (like Kaffeine) missing, this is not nice (but maybe this is a "intellectual property"
problem? Or hard and therefore slowing down - yes - the showcase)
- a lot of personalizations and tune and tweaks have been suppressed (claiming nobody used them)
But overall I did not find the applications this kind of disaster, so it promises well. It is just that everything is still unbelievably provisional, and nearly morphing. Every editions some very basic things change a lot (did I forget about KMail address list missing - but Kmail was pushed forward nevertheless for the sake of testing Akonadi integration and migration - yes - the showcase). Remember widgets management, same issue. By itself it is not bad to change, but I personally miss an understanding where the desktop wants to go esthetically, ergonomically and usability wise. Where it wants to go technically has been explained quite well.
And the more this goes on, the more I have the impression that the very team was only interested in the technical aspect but had no vision at all. Maybe they follow the views of the former chancellor Helmut Schmidt: who has visions should go to see a psychiatrist rotfl!. Seriously, it is IMO NOT just a problem of bugs and feature request, but a question of “quo vadis”. And I have some doubts that users will have a lot of say in it.
I hope this was not too much OT. And sorry for the long post. It is Friday. 
