But seriously, this seems to be checking for the antithesis to good code and I imagine a lot of programmers here are thinking 'you shouldn't need it!'. So sell it a bit more as your website doesn't.

On the code bases you've beta tested this on, do you really identify this kind of percentage of replication? Isn't that a sign the code is bad and should be abstracted or does this help in another way?

I'd want to see some real life examples on your site somewhere on some public repos, where are they!