Uh? The only radical mempool policy change proposal I see (removing incentive-incompatible filters entirely) are from Peter Todd and Ben Carman, but don't seem to have much following. Milder, more conservative proposals come from Luke Dashjr in Knots & disrespector-patch promoters in Core (extending to tapscript the very same datacarriersize policy that Core uses for pre-taproot txs) and Mark Erhardt (filtering out bare multisig in Core as well, as Knots already does since year). All minority views. Do you hear about anything else?