My impression of most of the user interfaces we encounter on the other hand is that they are build for the lowest common denominator and much of what the previous poster called "the science" is about how quickly the "on boarding" works, so how quickly someone can do a certain task when they are not familiar with the interface.

What seems to be never tested is, when people are very profficent, how long does it take them to do tasks. I understand why that is the case, it's much easier to do a quick study with some new users to test out a UI, but to design several UIs and then let people become very profficent with them first (possibly taking months) to then do a study comparing the interfaces is much more involved. So instead we extrapolate from the novice user studies to advanced users.