Pardon me, I see I have been downvoted for my observation, but am I wrong? If so, wouldn't you instead care to elicidate me?

Agreed

This is an awesome little tidbit of information.

@Rishat Nuriev:

see the fork: no meaningful difference between the three versions (500 tests on one input of size 500x500). The 3 versions perform about the same. Sometimes one is faster than the others, but not always the same one. You'd need something far more precise to actually see any consistent difference.

@Unnamed: you don't use lambda in there, so no overhead on this side (which is often what is slowing this kind of approach, I believe).

2 Usually true.
1 Who told you that? I know Python folks don't like lambdas, but I haven't heard anything about map. One function call is easier to read than an expression where a new variable is introduced for no reason.

Thanks for this, love these extra bits of knowledge dropped in the comments.