Did this highly scientific study contemplate the possibility that this is in part the result of people feeling like they're more justified in turning to piracy if a game is burdened with Denuvo?

Spoiler: It does not, so far as I can tell at first glance. It appears that the model is constructed entirely from DRM-crippled games that got cracked, and then then the estimate of how much revenue would be lost by going DRM-free from the start is extrapolated from that based on the assumption that it makes no difference. Maybe it's true, but the acknowledgement that it "can and often does cause problems, and some developers have chosen to avoid Denuvo altogether because it had such a negative impact on how well their game would run" sort of suggests otherwise.

https://abs.freemyip.com:84/share/_5WuM4QF — be careful following strange links you found on lemmy, but this appears to be the pdf.