Disagree here. Your main argument about the cons of BIP-444 is because it sets a censorship precedent. It doesn’t for two reasons: any amount of arbitrary data above 83 bytes is not improving monetary properties, it’s simply lazy dev approach. Bitcoin is either for monetary use or for everything else. You can’t have both. Second, limiting arbitrary data to 83 bytes in consensus can’t be used to censor monetary txs. This has been widely discussed and rebutted.