I'd say the exact opposite is true. Not wanting to implement the filter strategy relies on observations, accounts for human action, and takes the adversarial nazure of an open network into account. The filter strategy is mostly fueled by a mass call to policing action, an entirely moralistic argument, embedded in "if they would just" thinking.
Discussion
You are aware that the filters were basically present since ever. There is empirical evidence you can trace on the blockchain that the filters are working as designed, as 99% of all transactions adhere to them?
It's not the Knots people that want to change Bitcoin.
Knots is deploying policy rules to filter transactions that are widely used and broadcast. That is a radical take off from existing practice. Filters do something if miners use them, if not, we have two good examples now that they are virtually useless at prohibiting transactions from reaching miners, or communicating to miners that they should not include them in blockd.
Please prove your point with quantitative data. Otherwise it's pointless to discuss this. If filters are useless, we can just leave them in place, no reason to alienate the community about it.
Both full rbf and sub sat fee txs made it through what could well be called fully deployed filters. The op_return filter specifically created a perverse incentive that led to counterparty, omni, and stamps embedding data in outputs, which is strictly worse than if they could have used op_return instead. The rule should have been relaxed a decade ago in my opinion.
I stop engaging now, there are really no facts we could discuss about, I would like to see some data. Nobody thinks filters will prevent any non standard tx going through, but they rate limit them enough to not impact the intend of the network.