Ranked in these QuestionsQuestion Ranking
Pro Easy to maintain
Wayland has no drawing APIs. Instead, a Wayland client gets a DRM buffer handle, which is practically just a pointer to a graphics memory. Practically Wayland does not care how the client draws to that buffer, it only copies the client's buffers on the screen.
The removes a lot of complexity (because Wayland just pushes the complex stuff to the other layers of the stack) and by making the clients responsible for all the rendering, they can be smarter on how they do things like double-buffering for example.
Pro Simplifies the graphic stack
Wayland simplifies the graphics stack by trying to force everything through a GEM/DRM stack and straight into the kernel. Furthermore, it manages compositing itself.
Con XWayland handles popup windows poorly
Con No mechanisms to configure input
Tools like xinput and xmodmap that help customize keyboard and mouse input are incompatible with Wayland, have no corollary, and there is no clear roadmap for providing their functionality.
Con Little driver support
Most closed sourced drivers do not support the KMS/shared-GEM/shared-DRM technologies on which Wayland works. While this may be okay for open source purists, who only want to use graphic cards that have open source drivers available, it may not sit well with people who spend a lot of money for high-end graphic cards only to get some crappy 3D performance.
Although it should be noted that NVIDIA has declared that they will start supporting Weyland, it may take years before Weyland fully supports most high-end drivers.