The odd thing about that graph, to me, is the large number of estimates in the middle of the scale for P(simulation). I'd have expected more people to accept the simulation argument (and rate it very high) or reject it outright (and rate it very low). I guess we might be seeing uncertainty over whether or not to accept the argument, or some kind of computationally bounded simulation argument where the root universe can only handle a relatively small number of simulations. I don't quite think I buy the latter, though.