The problem with those sorts of studies is that they can make a computer system less productive *if* you have some experience in using the system. The problem them becomes very quickly *most* users have *some* experience of using the system and it becomes less productive.
For example most sensible people consider that using LaTeX to enter mathematical formulas is wildly more productive than using Word *provided* you have put in the effort required to learn how to use it.
Consequently there is no one best system. The problem is UI studies have a tendency to focus on making systems easy to use for beginners not more productive for those with some experience.