add benchmark numpy vs standard list in numpy.md

[ac812]

more notes related to this from Anton:

[ac812]

2) Some basics on good coding practice in algorithms.

Again, keeping the code as simple as possible, avoiding recomputing values multiple times and avoiding using more memory than required. Learning how to use timeit to explore which parts of your code are slow.

Something like here: https://www.geeksforgeeks.org/optimization-tips-python-code/