Because Wayland is being written primarily by former X developers who have pushed X to its limits but have no choice but to start from the ground up to get modern features such as tear-free drawing.
Strictly speaking that's not true, from the Wayland FAQ (emphasis mine):
Why not extend the X server?
Because for the first time we have a realistic chance of not having to do that. It's entirely possible to incorporate the buffer exchange and update models that Wayland is built on into X. However, we have an option here of pushing X out of the hotpath between clients and the hardware and making it a compatibility option.
I guess the main reason Wayland doesn't take so much flak is that it's obvious the mission scope has vastly changed from the 1980s display server to the 2015 display server. And it's main deficiencies are most visible in the markets where it's barely present (desktop) or has been replaced wholesale (Android), while the init system seems like you're changing a winning team, honestly when was the last time init scripts was a deal breaker for anything? It has a much more "nice-to-have" feel to it or at least fixing corner cases most people never noticed.