So, what's the "other side's" argument in this? Usually these opinionated changes come with some level-headed reasoning behind the changes. Or maybe not?
Gnome developers gave pretty thorough reasons why they made certain changes in UI and code design, or was I hallucinating that period of my life? I mean, just because one doesn't like designers' choices doesn't equate to "couldn't extract reasoning". Also, it's their project, and I hate that they ignored the user community it still fell under their authority to do it and let users decide if they would tolerate it or move to a different windowing system. Personally I just moved back to KDE and have stayed there, other than very limited resource systems where I use lxde or just straight terminal+tmux.
This is what I've been wondering here, as well. So far, the downsides (Change to an almost subconscious muscle-memory-task, added friction) of this are pretty apparent - what even are the upsides of this?
The new behavior doesn't make me want to invest effort in learning registers. A notification in the status bar which register just got filled would be all I want.
I imagine it's that reading user input through something other than the minibuffer is annoying, because if you have customized your input/editing keys, read-key generally won't reflect that.
Almost as it's just and only petty power trips.