Context for this?
The answer, for some of those questions, is like to be "none of them". Those that look problematic:
*We don't track self-reported language skills, because, well, userboxen are incredibly inconsistent between projects, and building up an accurate index would be a monumental effort. Wikidata would reduce the difficulty of this, but not to the point where it's either easy enough that it's no problem for someone to do in their spare time, or important enough that we've dedicated time to it.
*Tracking users between projects is something we're beard-stroking over at the moment because the CentralAuth infrastructure and db structure is the anti-pattern for "designing for researchers".