> Systematically well-done research with fake results or data is much rarer than just... lazy bad science.

Genuine mistakes, logical errors, and other oversights are even more common than that. For all the issues it has, peer review is quite good at catching the things that it's intended to catch.