I guess an extreme case of this popularity point is Roger Koenker who answered two questions on quantile regressions here and got 20 upvotes for them. The answers were very good of course but I would imagine that the name of such an authority in the field helped.

@Andy We have had other distinguished statisticians participate who were barely able to earn 2 upvotes per answer on average (which is well below the community norm), so name isn't everything: you have to contribute material that has recognizable value. (However, that in itself won't necessarily bring lots of votes on any particular post, as several have already observed here.)