Here’s a thought. If you’re developing something that relies on an open-source project, one of the big uncertainties is what is going to happen to the community in the future. Is this project going to wither away next year?
We could gather a lot of data on this. Mailing list use, code commits, blogposts, google hits, etc, etc, etc. Then we can try to see characteristics which are correlated with growing/declining projects, and develop a quantitative prediction of the likely future of a project.