Given that Mozilla and Google have already publicly objected to this proposal, I don't expect them to implement it. The W3C's word is not the law; no one is obligated to implement every specification they put forth.
It's simply a URI standard for crypto signatures. It provides no function except an address to something else. That's why Google is asking for a few "working" integrations to prove the theory.
Because someone goes to implement it and figures out the standard is missing something they need critically, they can modify the standard before it becomes a 1.0 standard.
Or, is it simply that Google, Mozilla and whoever else have to serve verification requests for their users?
Or is the whole joke in that none of this is figured out?