Note the point in the essay that statistical significance is meaningless if the model does not correspond to reality — which, in this case as in many, they very much do not.