These kinds of books teach you how to poke holes in statistical arguments. That’s useful, but it’s easier than using statistics in a positive, constructive fashion. And skewering a single line of argument is not the same as demonstrating the underlying claim is false.
Now the world is teeming with cheap skeptics who seem to know how to criticize papers - but they never actually learned statistics. All they’re doing is following a checklist to see if the paper adheres to overly rigid conventions. This explains the present misunderstandings about the absolute necessity of RCTs.