The thing is, even if performance was Gaussian, even if we run with the following 2 statements:

- IQ is gaussian

- IQ correllates well with performance

hiring practices would probably produce an employee population that went through some right-curve cutoff test, meaning most people would be much closer to the hiring threshold, with a few positive outliers.

For a given arbitrarily chosen values, you could massage the distribution and make it look Pareto, but I'd be hard pressed to come up with a reason why it makes rational sense.