Our opposite experiences just goes to show that UI qualities are subjective, and until someone releases scientific studies (none of this "many users" crap) that go against Microsoft's own user testing, I don't see why they should ditch their Metro experiment.