I don't think an outright ban would be acceptable at all or grounded in any kind of proportionality. It's one thing to use gambling as a guilt-by-association thing, but if gambling isn't outright illegal even in that somewhat fallacious interpretation an outright ban would be absurd.

Which is something I feel a lot of the people rallying against this practice often didn't think through, but hey.

I still disagree with your interpretation of that literature review.

This systematic literature review analyzes 190 empirical studies published between 2012 and 2023, revealing nuanced findings. Regarding compliance, 41% of studies reported high compliance levels, 29% low compliance, and 29% inconclusive results. For effectiveness in achieving regulatory goals, 44% found self-regulation effective, 33% ineffective, and 24% inconclusive.

Our review also finds that the presence of intermediaries such as industry associations, third-party auditors, and NGOs, along with certain types of state involvement, tends to enhance self-regulation outcomes.

That's less "it's a crapshoot" and more "it generally works, especially if there is an overisght body".

Which in this case there absolutely is, given that this all slots into pre-existing age ratings and content warnings. Your misgivings don't line up with the data you provide and don't line up with pre-existing analogous self-regulation.

I've seen nothing to suggest this is any more problematic than either other types of monetization or other types of content restriction, and the big differentiator between violent/sexual content and this seems to be whether the segment of the userbase that posts online likes it as a matter of creative opinion.