> proper language identification much thought, it seems the Working
> Group expects quite some change in the XHTML community. I do not see

Whilst I certainly agree with you with respect to the current mass 
market HTML authoring community, XHTML 2.0 is unlikely to be taken up
by the current HTML/XHTML 1.0 community as it is very much different from
the pseudo page description languages that they prefer.  There is some
risk of it being taken up for fashion/"looks good on CVs" reasons, but,
in general, I think most people who use it will take semantics a lot
more seriously than "web designers".

(The main risk to quality is going to come from companies who insist on
it for their own fashion reasons, especially if they use content management
systems and don't train employees in XHTML itself.)