In general, I do think that Darktable's UI could be better, and most likely it could be more efficient. His points are not wrong, but they are also not as fundamentally disastrous as the author makes them to be. I feel the damage that the author caused with his behavior is potentially worse for Darktable than the problems themselves.
In a collaborative project run by volunteers, the way one argues and the point one makes are united. Darktable is possible only if there is a group of people working on it. I've seen the author insulting fellow contributors in public forums, I've seen him using the most offensive possible language against people who were being polite to him, just because they disagreed. This is not how you steer an opensource project in the right direction. It is also useless to come and shit everywhere after the changes are done. It is the opensource equivalent of seagull management [1].
I think in a collaborative project, the way you state your point is as important as the value of the point you're making. If you're unable to state your point clearly and effectively, then you should work alone, as the author has now correctly decided to do. Continuing the shitstorm even after his collaboration with Darktable has terminated seems, to me, to be just the continuation of a useless feud.
I say all this as a happy Darktable user, and not a contributor at all. I also loved the author's contribution, and found his Youtube channel to be really informative and a fantastic learning source. Still, Darktable has continued after his departure, and the new release, which will come for Christmas, is really promising, with the inclusion of per-channel curves in the sigmoid module, similarly to Blender's AgX. Many other improvements are coming, and I don't feel the project is "crashing into the wall in slow-motion".
[1] https://en.wikipedia.org/wiki/Seagull_management