One thing I don't like from the start in this analysis is the assumption of a Gaussian distribution for power. If you're riding in the Tour de France, you're in

One thing I don't like from the start in this analysis is the assumption of a Gaussian distribution for power. If you're riding in the Tour de France, you're in the tail of a distribution of power generation of the human race. To assume it falls off symmetrically balanced in each direction from the mean of qualifying riders is wrong. The cutoff to even be able to ride in the Tour is far to the right on a Gaussian curve. I'm not saying there isn't a way to model what the poster is talking about, but I don't think this is it.