[lmm]

> the time is better spent on something that will produce demonstratable results, even if imperfectly applied.

How is formal proof not such a thing?