I think he published this video when I was starting to think about the same things, especially about working out, where it's more important to workout consistently with a decent form/weight than to spend hours reading up science and working out very little.

Most "benchmarks" or data that we have mean very little because of the context (or lack of) they are done.

I often read critics to benchmarks in here and how they are not "real life scenarios", seems like CS is not the only world where this happens