I follow, but your model seems so extreme as to be a self-characiture! :-)

What's the best evidence against it, or quick test that would be able to at least tell maximally naive model apart from maximally cynical model?

Do we really know totally nothing about that?