To be absolutely clear: I am NOT advocating for an increase in block size. I agree that that would be bad.
Releasing the OP_RETURN size limit does not increase block size. It doesnât even change whatâs valid on chain. It just means that the OP_RETURN portion of a transaction can be larger, and still transmitted by nodes before confirmation.
A person wanting to use OP_RETURNs to store data can already store as much as they want (within the block size limit of course). The consensus rules donât place a limit on this (to my knowledge). So you could already, today, make an enormous OP_RETURN and mine it yourself, or pay a miner out-of-band (like slipstream).
The hullabaloo over the OP_RETURN limit is about node *policy*, not consensus. It is the current policy of Bitcoin core nodes not to retransmit certain valid transactions. For example, zero-fee transactions are automatically discarded as spam.
In this vein, transactions with OP_RETURNs larger than 80 bytes are dropped by Bitcoin core nodes, even though theyâre legit transactions on the consensus level. The change being discussed is removing that *policy* limit.