Tech is full of examples of 'successor' technologies, that were aiming to provide a clean rewrite without legacy, which then got bogged down with supporting a bunch of corner cases and accumulated their own share of cruft and could be no longer be considered a cleaner alternative. All the while the majority of the userbase being stuck on the old platform because the new one is buggy and doesn't offer anything tangibly better.
Vulkan, various node replacements come to mind.
Wayland at this point has existed almost as long as X11, longer if you only count the Linux years, yet its still not quite there.
Wayland and its various implementations like KDE plasma are 90% ready. Now they just need the other 90% to move from alpha to product. I expect it'll take another decade or two.
The way things are going there might be a third 90% needed.