As regards the second bullet from the first key points list: there are many (many!) types of validity in empirical science, and I think that the quote is true only of some of them (construct validity, if I’m not mistaking). Am I wrong?