I think there are some interesting things that could be done with virtual reality in psychology research, but this particular paper strikes me as really problematic.
Different online games (or communities, if you want to be pedantic and say SL isn't a game) can have different cultural norms about what's acceptable behaviour. Second Life, in particular, tends to be more tolerant of nudity, open displays of sexuality, and unconventional sexuality (e.g. BDSM, furries...) than other online environments. Indeed, it's notorious for this. There's an obvious self-selection bias: the people who continue to hang around in Second Life are mostly the kind of people who enjoy the kind of things that go on in Second Life.
I really don't think you can generalise from Second Life to other online games, much less the real world. The result in the paper is just about a particular subculture.
(It might have been interesting to compare some of the "nudist beach" sims in SL --- I'd conjecture you'ld find a large number of gay men who enjoy looking at naked men --- which would skew your starts more towards male nudity, The point of the comparison being that the "gay bar" subculture has way more male nudity that the "star wars RPG subculture", with neither of them necessarily being typical of any wider group).