"N^2 is bad, except when it isn't" is my take on it.

In uni I maybe had one week of doing O-evaluations, it never came up and I never got interested. I've seen 20,000-word debates on Reddit on whether something is n, n2 or logn... and to my knowledge nobody ever learned anything from those.

In the workplace I've several times ran into the problem of "This is taking way too long," developing tools and methods to measure and drill down, then figure out if it can be improved upon or should be left as-is for now.

Honestly discussions and articles like this leave me absolutely terrified of interviews. My first and only technical interview was my future boss leaving me alone for 30 minutes in a room with a laptop "Code something you like yourself."

That man was brilliant.