Alright, I finished my first proper blog post about selective applicative functors!

I lay out the theoretical basis for them and how I claim they should model control flow with exclusive determined choice (which is something the existing formulation failed to capture)

https://blog.veritates.love/selective_applicatives_theoretical_basis.html

the key insight necessary to make it work for N-ary branching is to consider arrows (composable profunctors) instead of functors in isolation, because casing on an Either is about the domain not the codomain

the resulting structure is related to near-semirings, but in a different way than Alternative is (as that encodes nondeterministic choice)