[goodside]

These are all policy statements and intuition-based speculations. I don't doubt for an instant that many people have speculated about why diversity might be good for software. What I doubt the existence of are empirical studies showing that diversity benefits open-source software in some measurable way. The original article seemed to be under the impression that such studies are common, and that their interpretation is so thoroughly beyond dispute that the point needn't be considered further.

[rada]

Yes, it's better to base one's actions on empirical studies than on intuition. Yet we all have to put our pants on one leg a time, and it's far, far better to be out of the house in time for work than to sit there paralyzed and unable to make a choice between "left" and "right" simply due to lack of definitive statistical analysis.

As a tangenial but relevant point, there are empirical studies that show the inverse correlation between the number of programmers on a given project and code quality. Nevertheless, I have never seen anyone argue for hard limits on the number of contributors per OSS project. Methinks all the calls for hard data, while seemingly all about science, are really about emotion on your part.

(Sorry about the ad hominem - I am really not trying to attack you personally, I promise - I just don't know how to make my point without it).