Yes, many concepts are not easily quantifiable/mathematically provable, or if they are, would be too expensive to prove in a controlled, scientific study that could be done fairly. I also agree that is why belief systems are a driver for many cases. That said, I still offer some counter-arguments:

I don't believe that all you'd need is to show proof in order to convince someone who is invested in their belief system. This is discussed in the Mark Manson's "Everything is F*cked", where in attempting to show truths that threaten one's belief system, they may consider you as a hostile. I would love to believe this excludes computer programmers, but this can be observed in general human behaviour.

I don't preach my beliefs about SOLID principles and design patterns, anti-patterns, test automation, and source control and CI strategy for the purpose of superiority. I preach about them because programming can be hell without these things, and I've suffered alongside others working on buggy, fragile, unmaintainable codebases for years as well. I want less pain for those who have been in my place, but aren't aware there's another way.