I didn't go through and click all the links in the article, I was mainly interested in the problem-solving section. However, that one links to a Harvard Business Review study that drops this quote:
> Received wisdom is that the more diverse the teams in terms of age, ethnicity, and gender, the more creative and productive they are likely to be. But having run the execution exercise around the world more than 100 times over the last 12 years, we have found no correlation between this type of diversity and performance.
They go on to say that when they say "diverse teams solve problems faster", they mean cognitive diversity:
> Cognitive diversity has been defined as differences in perspective or information processing styles. It is not predicted by factors such as gender, ethnicity, or age.
Intuitively this makes much more sense than the race-based stuff I've been hearing about. You want people who can look at a problem from a different angle from you. It's sort of like having a team of specialists (one guy is good at UI, one at performance, one at tooling, etc.) except on the level of cognitive processes and not specific skills.
I believe that maybe these studies arose saying "diverse teams are better" and people misunderstood it to mean "racial diversity" and ran with it (kind of like the Hungarian notation fiasco).