Here is an analysis that cuts to the quick! Cui bono, indeed, if the basic data are already compromised? Your point about the illusion of statistical similarity, when comparing apples and oranges, is refreshingly insightful.