Doesn't seem compelling. After all, the root of the argument is about probabilities, but the argument then goes on to propose that we base our decisions on a single simulation of a result drawn from a weighted random distribution.