Larger OP_RETURNs may slightly reduce the friction for users to embed small amounts of data into the blockchain. Since OP_RETURN data is written into an output script, such a data payload is not subject to the witness discount. This means that larger OP_RETURNs would be significantly more expensive than inscriptions which are already happening.

After watching the NFT enthusiasts spend over $280M in transaction fees over the past 27 months, it seems to me that there are strong financial incentives for miners to facilitate such use. As the block subsidy is exponentially shrinking, these incentives will only grow. It will inherently be possible to embed data in the Bitcoin blockchain. We can make it a tiny bit harder by adding friction at the mempool level, which will incentivize mining pools to facilitate private relay and out-of-band submissions, benefiting the largest mining pools. We can make it significantly harder by restricting outputs to a small whitelist of payment output scripts at the consensus level, but that would neuter Bitcoin’s scripting system. The question boils down to what trade-offs we are willing to make and how much overhead there will be to inserting data—we cannot completely prevent it.

Bitcoin’s scripting system is what will allow us to build innovative scaling solutions and is necessary for the network to succeed in the long run. Mining decentralization was at its high point in 2017 and has been trending toward more centralization. Censorship resistance seems more important than fighting a few monkey pictures.

So, to me, it is preferable that a modicum of data be written to OP_RETURN outputs rather than to unspendable payment outputs that will linger in the UTXO set forever, and I don’t think it’s worth to make sacrifices regarding the scripting system or mining decentralization just to bring down the hammer on some Bitcoin users that happen to enjoy trading graffiti. Either way, the worst of the most recent colored coin and NFT wave appears to have passed. Feerates are down, and blockspace demand from runes and inscriptions appears to have jumped the shark: