Why is there so much interest in the Markov categories which are Kleisli categories for monads corresponding to distributions etc. but not much discussion of the E.M. categories? For example, the E.M. category of algebras for the finitely supported distribution monad - the so called finitary Giry...