This just isn't true, for the simplest possible reason: there's more preserved literature than there is preserved graffiti. It also isn't true for the slightly less simple reason that the literature covers a much wider range of topics.
What historians desire most is driven by what they don't have now. Cuneiform tablets are so numerous that they mostly just sit around untranslated. Would they be informative if we did translate them? Of course, but the manpower isn't there. Identical documents from 3rd-century Germany would be an epic, multiple-career-making find. Those would be translated immediately.
We value things the same way. Yahoo Answers is worthless to us because we have infinite amounts of similar material, so there's no cost to destroying this subset of it.
You could try to argue that Yahoo Answers has great future value, and we should therefore preserve it, but if we followed that advice, Yahoo Answers would have no future value, because so much content like it would have been preserved. This approach fails to be logically coherent.