And even for that, I think simply "taking turns" (where users share an editor session, can chat with each other, and can switch on sequentially who gets to actively edit) is enough for 99% of cases, and is not more difficult than mere single-person editing (since there are no conflicts).
To me, xi seems like an aspirational project, so this seems like the perfect place to take some time and design for these features up front.
Yes, but my comment also alludes to the opinion that "well working collaborative editing" is not a real problem people have, and a much easier (and easier to bolt on) use case, of consecutive (serialized with "locks") collaborative editing should cover most people's needs...
By all means, focus on creating a kick-ass collaborative editor, and add just the editing capabilities needed to make it good at collaborative editing.
It's the best advice in order to see any update.
There are plenty of programs that do some unique things very well, but fail on doing "everything that the mature alternatives do", so they fail to ever get mainstream traction themselves.
People want a complete solution that ALSO does X unique thing, if they are to drop their existing editors. Not something that they'll have to use alongside them for that special case.
(Joel on Software has written some nice posts about this idea, and why "minimal" competitors, who don't do "everything that the mature alternatives do" frequently fail, though I can't find the link right now)