Sociological data about trends in opinions, the opinions of newly tenured people, about the opinions of people in the newest branches of the field, etc. don't count as evidence?

Yes, that certainly counts as evidence if you're asking a yes-no question about whether one of these statements is true. But I think it's clear from the context that we're talking about evidence from real understanding of the matter, not just indirect evidence based on judging of what sorts of people believe what. Even though, of course, the latter can be perfectly valid evidence.