TBH I don't know if we can trust these findings. He goes into very little detail about how he manages to get his results. I'm not a fan of Denuvo, it is the bane of my existence when testing games, but I don't see how he could have tested this properly.
Do we really believe that he had these benchmark results just sitting there from before each DRM removal patch? That's convenient. How old were these legacy tests? Could other factors have affected his results, Windows updates perhaps?
Bulletstorm: Full Clip lost its Denuvo in 2017, so he sat on those results for a year? Life is strange also lots its DRM in 2017. Moto Racer 4 also lost its Denuvo in February 2018.
I don't see how he legitimately benchmarked these games. There is also the final test, which also looks bogus as his non-denuvo test doesn't even show his results, he just typed them and added them into the video. That's a little dodgy.
Perhaps this merits a little investigation of my own... some games have DRM on some platforms and no DRM on others. Perhaps there is testing to be done there.