Even less technical people feel comfortable discussing because lots of things in the past to draw parallels to --this has all happened before.

Expanding the mempool filters to also catch arbitrary data stored in tx witness data makes a lot of sense because it's how arbitrary data on Bitcoin has always been addressed. Expanding mempool filters doesn't break consensus.

The adversarial inscribers who spout " filters-dont-work-because-i-can-pay-miners" don't add anything to the conversation and forget that despite the increased fees from including spam, miners always "could" run the risk of getting their blocks orphaned when enough nodes on the network slow down the propagation of those blocks. The increased mining revenue will be short-lived.

It's not so much stopping inscriptions (can't be done), but discouraging those behaviors to an extent to get them to use something else to achieve their goals instead. No one is activating CTV to store jpegs when OP_Return could do the job just fine, and the disabled OP_CAT is a non-starter