Note that the statement is that the two people are identical other than their wealth, so that in this case the difference in probability is independent of how the original million was acquired. That's the point.
Of course they're not the same. The point of the experiment was to show that the differences don't need to be taken into account to replicate in a model the inequality we see in reality. The implication is that, while differences between people clearly exist, they aren't necessary to explain inequality.
