[BIP Proposal] Limit ScriptPubkey Size >= 520 Bytes Consensus.

Antoine Riard raises significant concerns regarding potential risks associated with a proposed rule change in the Bitcoin consensus, specifically highlighting the danger of confiscation risks for users utilizing bare multisignature setups. Riard points out that under the new proposal, both small and relatively larger multisig configurations could easily breach the newly proposed limits if uncompressed points are used, thus posing a real threat of confiscation. This issue is particularly alarming as bare multisig has been a practical choice for many, pushed further by an inadvertent P2SH script size limit which might have encouraged its use for larger policies.

To mitigate these risks, Riard suggests several possible adjustments to the proposal. One such measure includes applying the rule solely to transactions that do not feature an active nLocktime post-activation, or at least to one input with a height post-activation, potentially reducing the risk of confiscation significantly. Further, he proposes more conservative approaches like tracking the "maximum reorg height" (MRH) of every output to prevent transactions part of an unconfirmed chain predating the rule activation from being affected. However, this solution would require significant changes to current implementations, including additional data in each UTXO entry and a complete system resync, implying considerable costs.

Despite these suggestions for improvement, Riard remains skeptical about the overall value of the proposed rule change due to its inherent costs and the high confiscation risks it poses. He argues that while the intention behind the proposal is to clean up the UTXO set, the actual outcome might be counterproductive. The limitation could inadvertently increase the amount of non-prunable data within the network since entities could circumvent the restriction by creating multiple smaller outputs, thus exacerbating the problem the proposal aims to solve. Riard concludes that the proposal, even with potential modifications, likely does not justify the extensive costs and risks, suggesting that any benefits gained from keeping undesirable data out of the UTXO set are overshadowed by the negative implications and the technical and operational challenges it introduces.