Re: A bigger group...

I agree, I've made this proposal for CSSOM myself and perhaps it will get
some traction (though it will likely be through a different API).

Regardless, however, there are some extenuating parts so that is almost
unimportant by comparison at this point:  Tab Atkins from the CSS WG has
both agreed to take up numerous things which will make prollyfilling CSS
(and polyfilling) relevant and easier (http://www.xanthir.com/b4N80) and
has provided a fully compliant CSS parser in JS in the meantime but the OM
is slightly less than ideal for our cases IMO.  Simon Sapin has one in ...
python?  There are a few of them out there already - the trick is to see if
we can come up with one sensible OM which will let us do all sorts of
things (including pre-processing which wouldn't necessarily be solved by
adding features in CSSOM).  There are just too many cases where parsing CSS
and CSS-grammar compatible languages are useful to not have something like
that IMO - even after those other things are standardized.