A theoretical paper which shows how under certain mathematical definitions the title statement is true. I was hoping for some real-world ideas of how to measure "diverse" and "problem solver" and/or experimental data to back it up.
The best real world example I can think of is elitism, the idea that the smartest people are always the best for solving problems.
I like this article because it's related to the algorithm I came up with for handling group diversity with scaling issues. Still working on the implementation though.