This can already be done with an inscriptions-like construction. Furthermore, this can be done by any miner since there are no consensus rules governing arbitrary data. As above, this is not a new risk, and OP_RETURN does not make it worse. OP_RETURN actually makes the average block size smaller, and it reduces the validation required, so it's actually better for the network. Since OP_RETURN outputs are output data only, OP_RETURNs consume 4 weight units per physical byte of space consumed. This means that a block filled predominantly with OP_RETURNs will be smaller than one that isn't. Additionally, OP_RETURN outputs are provably unspendable. This means that nodes validating a transaction that contains OP_RETURN outputs do not need to consume RAM or disk I/O in order to insert the output into the UTXO set. instead, they skip it. This means validation of such transactions is marginally faster. The issue is that any kind of censorship leads to providing the technical ability to censor things which we do care about being censored. It's already possible to store arbitrary data, and more of it for less in fees, without increasing the OP_RETURN limit. However, there are some specific data storage techniques which people are (planning to) using which are harmful to the network. Specifically, there are people who wish to store more than 80 bytes of data in output data, primarily because they need to have some kind of commitment be broadcast publicly and widely. But since it is more than 80 bytes, they can't use an OP_RETURN and would instead use multiple Taproot outputs with the pubkeys as the data. Since those Taproot outputs are not provably unspendable, they will exist in the UTXO set forever, and that is damaging to the network in the long term. By increasing the OP_RETURN limit, we are encouraging them to use OP_RETURNs so that the data does not need to be stored in the UTXO set by all nodes for eternity.
This can already be done with an inscriptions-like construction. Furthermore, this can be done by any miner since there are no consensus rules governing arbitrary data. As above, this is not a new risk, and OP_RETURN does not make it worse. OP_RETURN actually makes the average block size smaller, and it reduces the validation required, so it's actually better for the network. Since OP_RETURN outputs are output data only, OP_RETURNs consume 4 weight units per physical byte of space consumed. This means that a block filled predominantly with OP_RETURNs will be smaller than one that isn't. Additionally, OP_RETURN outputs are provably unspendable. This means that nodes validating a transaction that contains OP_RETURN outputs do not need to consume RAM or disk I/O in order to insert the output into the UTXO set. instead, they skip it. This means validation of such transactions is marginally faster. The issue is that any kind of censorship leads to providing the technical ability to censor things which we do care about being censored. It's already possible to store arbitrary data, and more of it for less in fees, without increasing the OP_RETURN limit. However, there are some specific data storage techniques which people are (planning to) using which are harmful to the network. Specifically, there are people who wish to store more than 80 bytes of data in output data, primarily because they need to have some kind of commitment be broadcast publicly and widely. But since it is more than 80 bytes, they can't use an OP_RETURN and would instead use multiple Taproot outputs with the pubkeys as the data. Since those Taproot outputs are not provably unspendable, they will exist in the UTXO set forever, and that is damaging to the network in the long term. By increasing the OP_RETURN limit, we are encouraging them to use OP_RETURNs so that the data does not need to be stored in the UTXO set by all nodes for eternity.

This can already be done with an inscriptions-like construction. Furthermore, this can be done by any miner since there are no consensus rules governing arbitrary data. As above, this is not a new risk, and OP_RETURN does not make it worse. OP_RETURN actually makes the average block size smaller, and it reduces the validation required, so it's actually better for the network. Since OP_RETURN outputs are output data only, OP_RETURNs consume 4 weight units per physical byte of space consumed. This means that a block filled predominantly with OP_RETURNs will be smaller than one that isn't. Additionally, OP_RETURN outputs are provably unspendable. This means that nodes validating a transaction that contains OP_RETURN outputs do not need to consume RAM or disk I/O in order to insert the output into the UTXO set. instead, they skip it. This means validation of such transactions is marginally faster. The issue is that any kind of censorship leads to providing the technical ability to censor things which we do care about being censored. It's already possible to store arbitrary data, and more of it for less in fees, without increasing the OP_RETURN limit. However, there are some specific data storage techniques which people are (planning to) using which are harmful to the network. Specifically, there are people who wish to store more than 80 bytes of data in output data, primarily because they need to have some kind of commitment be broadcast publicly and widely. But since it is more than 80 bytes, they can't use an OP_RETURN and would instead use multiple Taproot outputs with the pubkeys as the data. Since those Taproot outputs are not provably unspendable, they will exist in the UTXO set forever, and that is damaging to the network in the long term. By increasing the OP_RETURN limit, we are encouraging them to use OP_RETURNs so that the data does not need to be stored in the UTXO set by all nodes for eternity.

This can already be done with an inscriptions-like construction. Furthermore, this can be done by any miner since there are no consensus rules governing arbitrary data. As above, this is not a new risk, and OP_RETURN does not make it worse. OP_RETURN actually makes the average block size smaller, and it reduces the validation required, so it's actually better for the network. Since OP_RETURN outputs are output data only, OP_RETURNs consume 4 weight units per physical byte of space consumed. This means that a block filled predominantly with OP_RETURNs will be smaller than one that isn't. Additionally, OP_RETURN outputs are provably unspendable. This means that nodes validating a transaction that contains OP_RETURN outputs do not need to consume RAM or disk I/O in order to insert the output into the UTXO set. instead, they skip it. This means validation of such transactions is marginally faster. The issue is that any kind of censorship leads to providing the technical ability to censor things which we do care about being censored. It's already possible to store arbitrary data, and more of it for less in fees, without increasing the OP_RETURN limit. However, there are some specific data storage techniques which people are (planning to) using which are harmful to the network. Specifically, there are people who wish to store more than 80 bytes of data in output data, primarily because they need to have some kind of commitment be broadcast publicly and widely. But since it is more than 80 bytes, they can't use an OP_RETURN and would instead use multiple Taproot outputs with the pubkeys as the data. Since those Taproot outputs are not provably unspendable, they will exist in the UTXO set forever, and that is damaging to the network in the long term. By increasing the OP_RETURN limit, we are encouraging them to use OP_RETURNs so that the data does not need to be stored in the UTXO set by all nodes for eternity.