It's true: we'd need a way to describe the tiny imperative steps. The benefit is that we DON'T need all the glue between them. Few languages and methodologies

It's true: we'd need a way to describe the tiny imperative steps. The benefit is that we DON'T need all the glue between them. Few languages and methodologies can deliver on this problem as well as Haskell can. Consider what my example is folding over and how those elements must connect. I'm not sure we could really do that monoidally. For a sequence of steps: probably. For opening doors and other conditional actions: we might need a more flexible strategy (I'm sure at the end though we'd call sequenceA on it!)