I believe the logic is:
1) Porn (crypto, gambling, etc.) transactions are significantly riskier than other transactions.
2) CC processing involves a long chain of intermediaries.
3) At every step of the chain there's strong pressure upstream for better rates, and strong pressure downstream to ensure the blended payment stream is safer.
So, eventually, someone in the chain gets told "sure, we can give you better rates, but only if you can improve the risk profile of your payment stream". And then they crunch the numbers, and decide it's worth it, knowing they'll lose some volume but hoping they can make it up with better margins. And then the people who churn find one of the remaining partners who hasn't adopted strict policies, and this repeats until all the "dodgy" payments are going through a high-fee chain that can't possibly afford to change their policies, and everyone else is going through a low-fee chain.
If you check CCBill, their "Blue" package that allows adult content runs 10.8% to 14.5%, with a $1,000 yearly "high risk registration fee". And even their normal plan (which doesn't allow adult content) seems to be 5.9% + $0.55 per transaction? Braintree is 2.9% + $0.30; Stripe is the same.
You can see why, eg, Stripe would be happy to leave CCBill to have a virtual monopily on adult transactions: There's not that many of them, they're so risky you'll need to charge absurdly high rates just to cover your costs, and just touching them at all will potentially taint your other transactions in the eyes of your upstream partners, driving up your rates. Conversely, you can see why CCBill will never kick the adult payments off their platform; it's the only thing they have to offer.
Disruptive startups work best when they can start with a niche and then grow from there, but the logic of the financial system makes that very difficult.