What I have found is if a majority of the network filters certain transactions, miners mining those transactions take on orphaning risk (due to worse compact relay) which can lead to them losing revenue.
This also makes it more expensive and involved to store data as Mara charges 2x the standard mempool fee, and methods like public keys cost more than inscriptions.
This does not stop data storage, but it does make it harder and more costly, eventually leaving them with nothing to do.
But filters don't and can't work for their stated goal.
The new wave of bitching is that a malicious actor could put CSAM on the chain leading to node runner arrests. All possible CSAM content is an infinite space and all possible encoding is an infinite space, so double impossible to filter preventatively. It would be hundreds or more blocks deep before the first arrests and we found out.
Anyone with that level of malice isn't going to be deterred by 2 sat a vbyte and possibly waiting 20 minutes instead of 10 on the 1 in 100 chance ocean hits the next block.
The XOR PR is a valid potential fix for that issue. Filters aren't. They are out here running a victory lap for a race they haven't even shown up for.
