This blog does pretty good job summarizing Naur's points. (My only quibble is equating mental models with theory in Naur's sense, but its an okay approximation). I definitely recommend people read the original paper. It's a flawed but beautiful paper.
Shameless self plug:
We just did an episode on Naur's Programming as Theory Building in the most recent episode of the Future of Coding Podcast.
I think all papers are flawed in their own way. So it is not meant as a slight. It is actually my favorite programming related paper. That said:
I think Naur was just a bit to vague and assumed background knowledge of Gilbert Ryle where most people don't have any. I think a lot of people finish reading the paper and don't realize what Naur's notion of theory is or how his radical conclusions follow.
Shameless self plug:
We just did an episode on Naur's Programming as Theory Building in the most recent episode of the Future of Coding Podcast.
https://futureofcoding.org/episodes/061