I think it's also worth mentioning that the other frequent response, "You need 100% of RDAs regardless of how many calories you consume", never has sources cited either.
If I may hazard a guess, I would assume that the actual answer is going to vary from vitamin to vitamin and from person to person. Assume a vitamin that every cell in the body needs a tiny bit of every day regardless of exercise levels. Two people with the same body types but different amounts of exercise would need the same amount (not proportional to calories). A very small person and a very large person, on the other hand, would need different amounts (proportional to calories).
And if we get into the more likely case of nonlinearly scaling requirements, both of the common responses are wrong. We (humanity) probably just don't know the answer yet, especially considering that RDAs are, from what I can tell, mostly determined by treating the body as a black box and seeing what amount of intake causes it to break. I don't know whether the scaling or non-scaling approximation is closer to ground truth.
I tend to lean towards scaling just because it's easier to formulate, it's the amount you get (in theory) from a balanced diet of conventional food, and it just sort of makes sense to me. But I don't actually know, so whatever. I'll probably be fine either way.