Arguably, by the time they implemented, they tested, at that point you're just consuming. So another way of handling this to contribute advancement at scale is

Arguably, by the time they implemented, they tested, at that point you're just consuming. So another way of handling this to contribute advancement at scale is "exercise left to the reader", and then we readers should publish too. R&D in big tech should contribute as academia does. We have our "Chief Scientists" and "Principal Engineers" titles that recognize, essentially, a tenure track, that should be on the merits of publishing code implementing new techniques.