Every single thing you've mentioned can be fixed with Xorg/X11.
(except subpixel antialiasing because that already existed for a long time, i'm not sure what you refer to)
Then why hasn't it been fixed?
In open source, those who make the code get to make the decisions--for better or for worse.
My personal opinion as to why X11 gets so few contributors--the build system is so terrible that nobody wants to touch it anymore. I suspect they would get a lot more contributors if they changed to something like meson/ninja which doesn't need to recompile the universe to be correct.
Are you referring to something else?
And I still see autogen.sh and Makefile.am goop in the top level of that repository.
However, I would be quite happy to be wrong.
I don't know, my guess is that people found making new stuff much more fun than fixing existing stuff.
Which is fine.
However that isn't the same as the existing stuff being unable to be fixed.
Cases in point: this project, https://github.com/const-me/Vrmac/ a non-trivial demo such as 3D teapot or 2D tiger.
Run that in desktop, and you’ll see occasional tearing.
Run the same demo on top of bare OS kernel (i.e. reboot to console), and it will work flawlessly on top of DRM/KMS, on the same Raspberry Pi4.
We had XV, RandR, antiaiasing and color managers with smooth video and no tearing since early 2000's. WTF are you talking about?
Except that X11 is NOT "mostly working" for a significant amount of modern usages.
Yes, if you happen to want to run your Terminal, Emacs, etc. remotely over a network connection, X11 is your huckleberry--but only because anything more complicated than a bitmap makes that a very difficult problem.
If you want HiDPI, subpixel anti-aliasing, color calibration, multiple resolutions on multiple displays, smooth video, no tearing, etc. X11 only works to a certain degree, some days, for some people, for some video cards. And it's not even clear how those should work over a network connection.
This is not new. After all, X11 critiques came in for a full chapter in the Unix Haters Handbook and that's 25 years old. Wayland is simply revisiting the same problems as X Server Extensions 30 years later (they aren't ubiquitous so nobody uses them so they aren't ubiquitous).
The problem is that people who use obscure features are VERY vocal about it while people who simply abandon your operating system because accelerated video doesn't work reliably are very quiet.
Maybe Wayland isn't the way forward, but X11 certainly hasn't been the way forward for decades.
The only advantage I can see is that if Wayland manages to tear out the X11-isms in everybody, what comes after Wayland will have a lot easier time of it.