I haven't seen many arguments worth addressing. They seem to rely on bearish long-term assumptions about monetary demand. So my advice to them is generally to sell it then?

The biggest technical concern I see is UTXO bloat, which the current op return drama doesn't contribute to. And the rapid UTXO bloat seems behind us now since it was based on a fad.

I view the node-level filtering for consensus transactions that entities are willing to pay for as being mostly performative. Node-level filtering works well for denial-of-service type attacks, not things that people are willing to pay for, since they only have to find one miner to accept it. If that type of filtering makes people feel better then I'm for it, but without consensus changes it's pretty trivial. It's good that people can fork and run their own node software, so it's a nice network-wide reminder of that option, but it doesn't meaningfully impact what gets into the timechain.