I agree with @Mia Chloe.

You can't prevent arbitrary data to be stored on the Bitcoin blockchain, there are way too many methods even if you remove OP_RETURN completely. The purpose of this proposed change has to do with incentives: if those using Bitcoin for data transactions switch to technologies which pollute the UTXO set (Stampchain and friends), then the cost of running a full node will increase.

This was also the reason why OP_RETURN was introduced at all. Because things like Ordinals (NFTs and tokens) already exist since 2013 or 2014, but the first variants of coloured coins and similar stuff were often polluting the UTXO set.

I however don't know if I really agree with this limit removal. It takes away a method to control the contents of the own mempool. I was quite happy with the current standardness rules, which should incentive OP_RETURN usage enough to prevent Stampchain-type mechanisms to become more common, but not enough to create incentives for additional bloat. So from my (advanced but not expert) point of view, NACK.