I mean i think the article is implying that. However i think the bigger thing is the correlation is misleading due to the sample being the long tail of github projects, which i dont think is representitive of "production" open source projects and certainly not software in general.