No, sorry, to be clear. I don't believe that "reduction to functions" is anything but a clever or perhaps illuminating toy and that (iv) is closed in the opposite way: it's certainly not true that "reduction to functions" is right.