Is psi an artifact of probability thresholds being inappropriately low over long series?

I agree, unfortunately proponents of psi choose to ignore negative trials. You can't sell books and pretend to be scientific unless you insist those tiny divergences are significant.