The author is himself imputing an absolutist interpretation of statistical significance. Statistical significance always has a p value (and an assumed distribut

The author is himself imputing an absolutist interpretation of statistical significance. Statistical significance always has a p value (and an assumed distribution!) associated with it. The permutation of statistical significances of insignificances the articles describe could well be true for various p values.