I believe the conclusion is basically true, but I also feel that sampling from the Github project showcase list induces a selection bias that is going to suppor

I believe the conclusion is basically true, but I also feel that sampling from the Github project showcase list induces a selection bias that is going to support it. The analysis would be more convincing if the authors had gained access to some enterprise codebases whose contribution graphs were more evenly distributed than their open source counterparts.