It's definitely possible some people do wrong stuff. I wouldn't trust people not to mess things up badly sometimes. But if you intend to criticize the whole discipline, you can't base your analysis only its failures.

If there isn't a reliable way of knowing whether its failures or its successes are the exception, I can see why this is a frustrating issue. Especially since both positions of for and against the approach ultimately have the same aim. The only real solution I see is solid data, but for something this complex and multifaceted that doesn't seem likely to happen any time soon.