Sure, it is fine to teach numerical linear algebra in Fortran - 15 years ago, I was learning with Fortran 77. Yes, there are many benefits of using a different language, even for its own sake. Or using a language without distractions - just to focus on the course, rather than one that can be used for building a website or so. Or without less "magic" than Python.

At the same time, I find most argument bizarre, other - harmful.

From the bizarre region, students asking "What is 'import numpy as np'?". A legit question. If you want to avoid all technicalities, pen and paper is the right approach. All in all, they need to run code somehow and there will be questions like "What is 'implicit none'?"

If they are not asking such question, is is not because they focus some much on linear algebra. It's because they are lost - or at very least - have lost interest.

From ones (in my opinion) actively harmful:

> It’s about teaching the basics of scientific computing to engineering students with a limited programming experience.

This is a big red flag. If they have little background in programming, teach good standards of Python. Otherwise they will get worst patterns, and think that GOTO is the preferred way of coding.

If you think why a lot of academic code looks like mess, one of the reason is using archaic tools and techniques as a standard.

Also - if your focus in linear algebra and avoid magic, you CAN do it without any numpy, just Python lists.