Many are excited about Wayland and Mir, but still these new display systems present new demands on the Linux desktop stack in terms of graphics drivers and acceleration, there's issues with non-Linux operating systems, legacy X applications continue to need to be supported, and there will be enterprise Linux distributions that still need to maintain X support for years. There's also some that are no longer excited about Wayland/Mir and are excited for the X.Org Server with DRI3 improvements. Simply put, X12 still stands chances of materializing.
The X11 protocol is now 26 years old but there hasn't been an "X12" yet since the protocol has managed to be extended many times and new additions made but without breaking the original core protocol. Many extensions have been introduced to X11 over the past nearly three decades like X RENDER and RandR while many parts of the protocol are simply not used (e.g. X server-side fonts) and most tool-kits and other applications now bypass large portions of the original X11 capabilities.
Back in 2010 there was some talk about X12, but it's more of a discussion to be had over a couple beers or in our forums than some serious, concrete planning. With Ubuntu delaying Mir on the desktop and Wayland not yet taking center stage on other distributions (but hopefully with Fedora 21, et al!), it's interesting to think about how long the X.Org Server will be in the spotlight and whether there will be an X12.
For quite some time various ideas have been collected about what an X12 protocol would look like and they are expressed on this X.Org Wiki page. Among the changes sought for an X12 would be to have an object model, fine-grained events, assume compositing by default, core rendering inside X has become largely useless, there's encoding bugs present in X11, the X extension space and XIDs need to be larger, SVG cursors would be helpful, XGE by default, moving X11 "fixes" into the core protocol, implementing ICCCM suggestions, and various other low-level changes. X12 would likely allow for more parallelism to work better on massively parallel systems.
The expressed "good stuff" about X11 right now is the network transparency, multiple platform support, handling of modern graphics hardware and rendering, the frame-buffer concept is still useful, and being as efficient as possible as a low-level protocol.
Well, what do you think about an X12 or will Wayland and Mir simply takeover? Let us know in the forums!