The spam issue does need addressing.
BIP 444 is like trying to do so as ineffectively as possible while also being sure to do plenty of collateral damage for good measure.
I'm resting somewhat easier knowing libbitcoin v4 is just around the corner and with no utxo set maintained and a different model of mempool it does do a bit of an end run around the damage caused by spam for many.
That said, it's not fit to be the majority implementation, and at some point we need to do something to defend the utxo set. I think it'll need to be both simpler and more effective than 444 -- something perhaps like removing the segwit discount and incentivizing it going somewhere else, be that op_return or preferably off chain entirely. But I don't remotely think the effective solution has been proposed yet, and the tribalism is exhausting.