I think the actual problem is that a lot of the functional programming discipline is about writing code to be understood by machines (eg: compilers, theorem pro

I think the actual problem is that a lot of the functional programming discipline is about writing code to be understood by machines (eg: compilers, theorem provers, etc) in ways that can trade off against the convenience of people reading and writing it. When I have to write a program that computes functions of programs, I always reach for my FP knowledge, first thing, and start using the more "theoretical" branches quite quickly. But program analysis and synthesis are different actual endeavors from writing some bog-standard, everyday code.