They provide a lot of references in the article to other research that make similar claims from both proprietary and much larger open source projects. At the ve

They provide a lot of references in the article to other research that make similar claims from both proprietary and much larger open source projects. At the very least the "Hero model" seems to be as prevalent outside open source as it is in open source. The purpose of the paper was to test and infer from a larger dataset, which open source provides more easily then proprietary. That being said, it really seems like the research was highly guided by an agenda, and I'm not sure how much one can take from this research.