I'm very comfortable saying that Sjors is intellectually lazy outside of the narrow domain that he wants to be working "productively" in. He just doesn't want to consider the ideas that:

• once you agree that non-monetary transactions are definitionally spam (he doesn't, meaning there is evidence of cultural shift within the core developer community) then no methods of reducing those transactions entering anyone's memepool is akin to censorship
• it is not a problem that spam reduction and ongoing cat and mouse game that is never "won" even if he specifically doesn't want to work on it directly, as long as he doesn't prevent others from doing so
• it is not a problem that people "look" at their memepool by auto-filtering out transactions based upon certain criteria that they may set subjectively, choosing which to ignore entirely or which they don't wish to relay, therefore signalling to miners what kinds of transactions are valid to them at the policy layer. There is no reason to believe that this leads to censorship due to orthogonality. To reduce bitcoins censorship resistance almost all full nodes to coordinate in order to achieve an effective sybil attack by ignoring transactions that meet certain criteria beyond those that people overwhelmingly agree are non-monetary.
• The argument that all transactions that are consensus valid should be relayed and that the network only really needs blocksonly full nodes in large quantity equally applies to bitcoin node implementations. Bitcoin implementations that add filters are consensus valid, and thus in principle and in practice the policy layer governed by node runners policing what is and is not an acceptable transaction to relay on a one-node-one-vote model of consensus has always existed.

Bitcoiners need to understand the classical legal concepts of malum in se (bad in itself, e.g. murder, or in our case "invalid transactions") and malum prohibitum (bad by regulation, e.g. cheating on taxes, or in our case spam).

Malum prohibitum is the regulatory layer, which are constantly refined and imbued with the values that dominate the current context.

With this current episode in Bitcoin's history, even if we have a miner and temporary setback where obscene and illegal content gets added to the chain, it is unlikely to result in a hard fork, but get interpreted as the kind of original sin that stains the network and provokes the flowering of hundreds of alternative implementations, forks of Core, that will succeed at regulating in spite of what the current crop of central planners think.

Just as it is impossible to stop spammers, it is equally impossible to stop nodes regulating the the network at the policy layer.