That's because the Kicad developers have a suggestion that goes along the lines of: "If there is a question about how to implement a feature, look at how the leading Maker favorite package does it, and do the opposite."
Good one!
Thing is, as I understood it, KiCad was once developed by a single guy who, if I'm not mistaken, was a teacher. Impressive work for a single developer but... the context says it all, pretty much. The key GUI features come from the original ideas, especially in the schematic editor which is the part that has been the least reworked of all. Anyone that had to deal with software written by university teachers will probably know where I'm heading to.
I've taken a look at the source code, although not extensively, but changing the GUI substantially would probably require a complete rewrite.
To be fair, they have done a pretty decent job with the layout editor's overhaul.
Everyone who is interested in the future of Kicad should follow the developers' email listserv.
There are major changes afoot to the schematic editor, in terms of the library system (to make it more like the layout), to harmonize the hotkeys and such so they are consistent across all of the applications, and moving it all to OpenGL/accelerated graphics. So, yeah, substantial rewrites, and they're actually aware of all of the complaints.
Trying to support three operating systems with a somewhat cranky cross-platform GUI, and moving forward while not breaking existing designs, is quite the challenge.
And if you have a dozen people in the room and ask them what they want in the user interface, you'll get two dozen somewhat incompatible answers.
I've built a half-dozen boards in Kicad, and I use Altium at the day job, and I've used Orcad and Accel and PADS in the past. They all have their user-interface and library quirks. And you can design boards in all of them.