Leaving aside my distaste for <meta> hacks for a moment...

In what way will this ease "batch validation" and what's the relevance of "unknown content" to the issue?

E.g. for http://www.w3.org/Consortium/Member/List I am not interested how valid 
the redirection pages the document links to are, but rather how valid the final 
destination is. Is that not obvious?

I don't see a <meta> refresh on that document.

While you might not care if a redirecting page is valid or not, others could
well be. If for some reason I were to use a meta refresh (which is very
unlikely, http responses are much cleaner) then I would be very interested in
knowing if the document was valid or not.

There are also those circumstances where a meta refresh is used, but not to
redirect a moved page. For example, a slide show with a refresh endlessly
cycling to the next page in looped sequence every 20 seconds.

Note that I was talking about the documents linked from that document and that 
I was not suggesting to do that by default, just that the Validator should be 
able to do that if I want it to do that.

Yeah, probably not a bad idea to implement at some point.

Leaving Severity=Enhancement and reassigning default component owner
(resetting Status to NEW).