I disagree. There are plenty of people who I would consider excellent engineers, both in software field (Linus Torvalds) and outside it, but they are not doing

I disagree. There are plenty of people who I would consider excellent engineers, both in software field (Linus Torvalds) and outside it, but they are not doing any mathematical modelling (except in the sense I outline in the other comment). They are just using lots of heuristics, experience and common sense. But in more mature fields, like mechanical or electrical engineering, such approach is more historical, since the theory and modelling is very well developed. Still, to restrict engineering only to that sounds very wrong from the historical perspective.