Is this objectively worse than the current practice of removing or hiding every visual cue and affordance, leaving the user to guess what they're supposed to be doing?