So the removal of OP_RETURN size limit will allow for larger data payloads to be embedded into Bitcoin transaction, which potentally could lead to significant increase in the blockchain size, unlike witness data wich can be pruned in segwit tx's the OP_RETURN data is stored in the UTOX set which will lead to permanent bloat as the blockchain grows. Increaing computational requrements storeage and bandwith to run a full node.

This could price out individuals and smaller users from running full nodes dew to increased cost or higher hardware cost. The effect of this could lear to fewer nodes and have a impact on the decentralization of the network.  A less decentralised network is more vulnerable to censorship, collusion or controll by a small set of well-resouced bad actors. Miners or corporations. 

The pull request acknowedges that some parties are already store data in unspendable output adding bloat to the UTXO set, but also argue that OP_RETURN is less harmfull of a alternative.

However, removing the limits entirely risks exacerbation of the problem rather than a mitigation of it.

Many will claim that data anchoring is inevitable and that OP_RETURN is more efficient way to handle it compared to fake addresses or unspendable outputs.  But the argument that the market for block space will naturally regulate use via fees.  This assumes miners prioritize long term netork health over the short term fee revenue which isn't guaranteed, especially given the pull request not that major miners already bypass these limits for profit.  As many see it Bitcoin core's streanth lies in its simplicity and stablilty which minimize attack surface and ensure long term reliablity.  Removing OP_RETURN limits and associated configuration option such as -datacorrier and -datacarriersize introduce new code paths and behaviours than mist be tested and maintained.

Larger OP_RETURN outputs could amplify existing attack vectors, such as mempool spam or flood-and-loot attacks, as noted by commenter Brazy Development in the pull request.

An attacker could flood the mempool with large OP_RETURN transactions, increasing memory and bandwidth demands on nodes. While nodes have resource management tools (maxmempool size, minimum relay fees ect), these are tuned for typical transaction patterns.  Massive OP_RETURN payloads could push smaller nodes to their limits, delaying transaction processing or causing nodes to crash. Attacks could also cause some disruption the network’s reliability, increase transaction fees, and disadvantage smaller miners who rely on efficient block propagation. The pull request notes that slow block propagation caused by non-relayed transactions getting mined already benefits larger miners, and removing OP_RETURN limits could exacerbate this by enabling more data-heavy transactions.

Proponents such as James Lopp are giving the argumentthat the market for block space and existing node defenses like fee based eviction mitigate these risks. It's being claimed attackers would face high economic costs. However, a well funded attacker like a state actor or competitor chain could absorb these costs to destabilize Bitcoin, and smaller nodes would bear the brunt.  The frustration everyone feels with developers “playing with people’s money” by adding complexity is a common sentiment among Bitcoin maximalists. The pull request’s proponents, like petertodd, argue that OP_RETURN limits are already bypassed and cause harm (UTXO set bloat from fake addresses), but their solution removing limits entirely feels like a capitulation to non monetary use cases rather than a defense of Bitcoin’s core principles.

Why we are not focused on more payment solutions and being the money anymore is beyond me.

What once was is no longer and this could be the very turning point many had the fear would come.

Don't forget where we came from and don't make changes that could damage the nature of the network.

But hey. What do I know.