> This is the kind of worldview that comes from years of navel-gazing in a therapist's office. "Why do I beat my kids? ...Because my father beat me! Now that I understand why I do it, I can stop beating my kids..." Or, you could just stop beating your kids.

Which of these approaches is more likely to result in someone changing their behaviour? Is there any evidence that advice to “just stop X” is as or more effective than causally aligned methods?

What an unusual attitude for an engineer to have. I suppose you fix bugs by “just not writing bugs in the first place”, too.