Curry-Howard is central to how you think about programming vs. Curry-Howard is almost completely irrelevant to how you program?
But maybe that disagreement goes away if I restate it slightly: Curry-Howard is central to how you think about programming vs. Curry-Howard is almost completely irrelevant to how you actually write programs. That is, if I am actually in the process of writing a program, I don't think about Curry-Howard very much, even if I'm working with the type system in Haskell. If I think about programming at a highly abstract level, then, OK, Curry-Howard might be considered a central insight.
But maybe that disagreement goes away if I restate it slightly: Curry-Howard is central to how you think about programming vs. Curry-Howard is almost completely irrelevant to how you actually write programs. That is, if I am actually in the process of writing a program, I don't think about Curry-Howard very much, even if I'm working with the type system in Haskell. If I think about programming at a highly abstract level, then, OK, Curry-Howard might be considered a central insight.