That's a common misconception. Indeed, Citrea is frequently writing a state diff and a corresponding proof into the blockchain, but they are using inscriptions for that. Those OP_RETURN transactions are something different. They are used only in case of a dispute during pegout for watchtowers to provide a chainstate proof for a heavier chain than the one the operator claimed to be the canonical chain.

These OP_RETURN transactions likely never hit the chain because there are strong disincentives for operators to make any invalid claims. Thus, ironically, lifting the OP_RETURN limit hardly has any effect in practice. It barely effects the Citrea protocol and it likely doesn't effect what kinds of transactions we see in the chain.