While making constructive mathematics rigorous was Martin-Löfs original goal, he was quite early aware of the applications to computer science. See for instance his 1979ish paper «Constructive Mathematics and Computer Programming»[0]
I've used Haskell for 18 months in a professional setting. I prefer C++. Grass is always greener - having endless abstractions to be mathsy becomes a real pain when you're trying to build real products.
[0]: 1982 version: https://raw.githubusercontent.com/michaelt/martin-lof/master...