I see the problem that v30 is trying to address, but I completely disagree with the solution they chose. The concern is understandable: if users can bypass node relay rules by sending large OP_RETURN transactions directly to miners, that creates centralization pressure. Miners could start offering private submission channels for large data transactions. Over time, that weakens the public mempool and gives mining pools more gatekeeping power. After recognizing this issue, Core likely looked at different ways to handle it. In the end, they decided to align relay policy with consensus and completely remove the OP_RETURN size limit at the policy level. The idea was straightforward: if the standard relay path allows what consensus already allows, there is no incentive to route transactions around the network. Up to that point, I can follow the reasoning. But that is not the only possible solution. Why not go in the opposite direction? Instead of loosening relay rules, why not tighten consensus to 80 bytes? That would have: • Closed the bypass vector completely • Removed the incentive for direct miner submission • Kept the attack surface smaller • Avoided normalizing larger data embedding • Reduced legal and reputational risks • Made large-scale spam more expensive, since it would need to be split into multiple transactions In short, instead of expanding what is permitted at the policy layer, consensus could have been made stricter. So the real question is: Why was this option not seriously debated? Were there strong technical reasons against it, or was it simply not the direction they wanted to take?