The best engineer I've ever known ended up working for years on optimizing ad space auction time by micro seconds.
a) don't care
b) were desperate enough at the time, then, like that damn videogame, it sucked him in
it's too easy to get carried away by sheer technical complexity of optimization tasks, even if you are optimizing for bad.
Is the cleaner regularly removing poop stains from the personal toilet of a big and rich Google shareholder more useful than the qualified Google engineer working hard so a big number is very slightly bigger on one the shareholder’s list of numbers? I think the cleaner has more impact.