Permanently Deleted
Def not what you're asking, but there's these
https://github.com/trending
https://gitlab.com/explore/projects/trending
Probably have the same problem as pkgstats
Really good =/= popular software, so...
https://github.com/luong-komorebi/Awesome-Linux-Software
Removing any package that starts with lib should get rid of most of them. If your dataset has a libs section like debian, remove those too. Then you can manually review any stragglers.