I think if I were to try to make any DE work the way I'd like it to it would have to be my full-time job for a while for there to be any hope of it being finished any time this decade. The list of changes is huge irrespective of the DE used as a starting point, and the changes large enough that upstream probably wouldn't want them which brings the overhead of maintaining a fork.

Unfortunately, though it's a goal I'm working on, I'm not going to be in any shape to quit my job and start full-time tinkering on FOSS any time soon.