And what's the border to that? Alright, softfork says that more than 1 OP_RETURN is invalid. Great. Now, people don't use OP_RETURN at all. They store their data in chunks of 160-bit addresses, and send 0 coins to multiple of these addresses. What next? Invalidate transactions which spend 0 coins? Alright, they then spend 1 sat for each. You've probably guessed how this goes. At some point, they become indistinguishable from regular transactions.
You are forgetting that history has already debunked your speculation. When the limit was placed on OP_RETURN it didn't cause any issues nor did it push people into using workarounds. Basically the limit (ie. the standard rule) greatly and successfully reduced the amount of abuse.
That's exactly what we've needed from day one and I even said it from day one too: Ordinals transactions need to become non-standard. That may not stop the attack entirely but it would reduce the spam by something like 90%.
I'm going to argue the Ordinal transactions aren't monetary. They all pay miners, don't they? Transaction is information. The only way to distinguish transactions is based on what they pay the miner. If you introduce stuff like "Ordinal-like txs are excluded from the network", you're just digging your own hole.
I'm impressed Bitcoin experts from this board are yet to realize that the real attack is the attempt to invalidating transactions which they don't think "they are worth it".
You know why I keep using OP_RETURN example? Because it surprises me that nobody ever complained about that limit and a ton of other similar limits that existed in Bitcoin for many years as standard rules rejecting many types of "spam" transactions all this time. Nobody has been calling them "digging your own hole" all these years when we were rejecting them either.
So why is Ordinals Attack different all of a sudden?!!
The details of the attack is different but the nature of it is the same. It is injecting arbitrary data into the chain or in other words it is [ab]using bitcoin blockchain as a cloud storage.
Do you really believe that the miners will care about standardness when it comes to $50 million? Additional Ordinal software can be developed that sends transaction directly to mining pools, and by the way, that's worse because now we have less accurate estimation of the current transaction fees.
This is why I warned at the start of this attack that it should be prevented soon. As the scam market grows it attracts more people and more incentive to find workarounds for any kind of "soft" preventive measures.
In early days when this was a scam that nobody had even heard of, nobody would have bothered using an alternative risky software to buy a garbage that has no value.
In any case I still don't think it is too late. If the majority of nodes start rejecting this type of spam attack (like some of us do) they won't reach a mining pool to be mined.