English — KDE4 overtaking GNOME in terms of usability?
f it hasn't happened already it seems poised to happen sooner or later, albeit usability is to some extent in the eyes of the beholder. My basis of this expectation is personal experience of KDE4 improvements thus far and observation of future trends for both platforms.
GNOME is facing a fairly major shift next year with GNOME 3.0 and the introduction of GNOME Shell, possibly comparable to the shift KDE had with its 4.0 release. I could argue that the GNOME shift would be even more dramatic, thus taking an even larger risk. Fundamentally, in terms of user interaction with the system KDE4 actually didn't change all that much. It remained a somewhat Windows-like interface with a typical panel-desktop paradigm. It has "merely" upgraded it and rebuilt the gears behind the scenes which is where the most revolutionary changes and most potential seems to lie.
Perhaps the biggest change visible to the user is making everything into a widget, but the default and typical configuration of those widgets still resembles a typical paradigm, and that seems to sit well with the users. At least it does for me. I'm increasingly finding it more natural to have a single panel at the bottom with all of the stuff I need immediately available than to alternate between upper and lower panel.
GNOME 3.0 on the other hand promises to completely do away with the typical main menu currently resembling Mac OS X and replace it by something that probably wont be very familiar to neither Windows nor OS X users. Upon pressing of the "activity" button an entire screen transforms. It's not just another menu, it's a whole new "mode" of operation. It's yet to be seen how are users to respond to this as they begin to test it in day-to-day use, but I am somewhat skeptical and tend to side with some of the Bruce Byfield's commentary. Furthermore, as he points out, it doesn't appear that the GNOME Shell is getting enough opportunity to be tested by users for valuable feedback. This increases chances of GNOME 3.0 becoming a similar scenario to KDE4.0.