Proving a negative, eh? Good luck with that. But I'm sure with a carefully chosen game, a carefully chosen setup, and a carefully chosen independent benchmarker, then it won't have any problems.
Besides which, if we're linking PCGamer, here's their Resi 8 Village comparison between the Denuvo and the cracked version, showing the Denuvo version stuttering like crazy compared to the butter smooth cracked version. Probably the same maximum framerate and the same median framerate, and I wouldn't be shocked if the cracked version had a stutter in it, so the minimum framerate would be the same too.
In this case, we don't know all the variables: DRM may not be the only difference between the retail and cracked copies of Village. As Leadbetter points out, though, if DRM is not to blame for the stuttering, the only remaining conclusion is that the cracked version both bypasses the DRM and fixes a performance issue unrelated to DRM. That's possible, but I think it's safe for us to reject 'oops, accidentally fixed the stuttering' as the explanation here (and it doesn't really reflect better on the game)
It’s not too implausible to me that the developers usually rest without Denuvo, and then have imperfect methods of compiling Denuvo into the final product. There’s obviously some explanation for why some games do not increase in performance.
Denuvo is getting bold lately now that no one left to crack their DRM (unless there are new groups that said they'll work on cracking denuvo recently?).