Monte Carlo simulation doesn't generate all classical probability distributions though. Attempting to do that would be nuts. Monte Carlo simulation only samples from a single, fixed distribution (or maybe a small number of distribution).
I'm not super convinced by the argument based on number of parameters either, but your analogy doesn't refute it at all.
I'm not super convinced by the argument based on number of parameters either, but your analogy doesn't refute it at all.