I've seen this before
Usually spam that uses random junk to thwart Bayesian filters (beating the threshold).
Does it work? Dunno, but it seems like the wrong approach. It's still looks like a weak position of defence, rather than the frontal assault that's needed. At best it's a tit-for-tat escalation which those with the most money will always win, or at least always be out ahead.
The solution is pretty obvious: stop visiting abusive websites. That includes facilitators that also link to those abusive websites (affiliate spam). We already have DNSBLs like Spamhous and SORBS for email, why not have a similar system for websites? And I don't mean just blocking popups and ads, a la AdBlock, I mean block the whole damned site.
Once they see their traffic (and revenue) plummet, they'll soon get the message.
The problem is, as ever, is convincing Joe Public to care, certainly enough that he'd voluntarily block access to his beloved Fsckbook, for example.
Sadly, for that reason alone, any measure that requires voluntary participation is doomed to failure.