Completely agree.

Python users shouldn't run standard loops in any scientific computing. Only valid examples are looping over a set of http parameters in a web app...

In which case, the computation will definitely not be the bottleneck.