In past attempts to discuss this, the conversation was typically driven into "if it can't be perfect, why even try" kinds of rhetoric. The primary criticism being that you can still use some sort of vision AI/robot to cheat to some degree. My counterargument is that there exist systems beyond deterministic DRM approaches that can be blended in to make it effectively perfect for all practical purposes. Statistical analysis (aka you might be cheating) combined with dynamic active canaries (i.e. fake, 2-frame targets you bait the bot into aiming at) can solve the problem really well. You could put statistical bounds around the likelihood of fair play that look like IBM mainframe uptime figures.
The only other reasons I'd expect downvotes: Extreme principled ideologies (aka no DRM ever no matter what I don't care). Or, those who are actually cheating at video games and can't imagine a world where they'd be forced to bust their ass to get ahead at something.