The only way the utxo set is less stressed under a bip-110 future is if you assume the advocates are correct and spammers will leave the chain because we have signaled that they are not wanted. They are assuming things will operate in an ideal way. They are fools. All that will happen is that some amount of spam that used to take 1 transaction will take more and each extra transaction is utxo bloat. Spammers will simple use p2sh-wrap or multisig keys to embed data. You turn a utxo efficient system into a fragmented nightmare and accelerate centralizing forces. Gun to my head I’m picking larger a hard drive over having to get all new hardware to up my ram.

Learning here... So pardon my ignorance. If the argument is that spammers are adversarial and their activities damage the network with fragmented utxo's, then why would we suspect they would just use op-return? If they're adversarial and looking to damage the network, then leaving a door open for them doesn't mean they'll stop breaking windows. And second thing I'm not clear on, one of the major arguments I heard is that this is a policy change and not a consensus change. That's true, but in the same sentence, it's often said that nodes are irrelevant unless they're "economic" nodes. If nodes are irrelevant and only economic nodes matter, then why remove a policy that is configurable to the user anyway? Economic nodes can just change their own policies to allow the bloated op-return and never look back. From that perspective it seems that the move is a "soft" consensus change. Removing user configuration options forces a change of network behavior without changing consensus.