Systemd is creating the same kind of monolith monoculture that Xorg represented. Wayland is far more modular.
Regardless of your engineering preferences, rejecting change is the main reason to object to both.
Not sure I agree here, assuming you mean "... than X11". With Wayland, you put your display code, input-handling code, compositor code, session-handling code, and window-management code all in the same process. (Though there is a Wayland protocol being worked on to allow moving the WM bits out-of-process.)
With X11, display and input-handling are in the X server, and all those other functions can be in other processes, communicating over standard interfaces.
That's an implementation detail. You can absolutely separate one out from the other and do IPC - it just doesn't make much sense to do so for most of these.
The only one where I see it making sense is the window manager, which can simply be an extension/plugin either in a scripting language or in wasm or whatever.
I do dislike System D for two reasons. One is exactly because it s a monolith and, in effect, an extension of the OS. The other is the attitude of the developers which becomes very evident if you browser the issues.
Our OpenBSD packager has already said in our Matrix channel that he'll be testing here and there in order to keep me honest ;)
And in fact has been: https://github.com/wayland-transpositor/wprs
Only in reductio ad absurdum.