Feels relevant. OP_RETURN is configurable via bicoinds -datacarriersize flag. It is not technically a consensus rule.
Samurai uses OP_RETURN for both its bip47 paynym and Whirlpool Tx0 construction. The 46 bytes in the whirlpool OP_RETURN contains info to allow the Whirlpool server to verify that a user has paid their fees. Samurai probably does it this way because the limit was increased to 80 bytes in 2014 and hasn't really been changed since.
Although I probably should add that technically knots has had this limitation of 42 bytes Almost since it's inception, so the above image is technically wrong.
TDDR (Too Dev Didn't Read):
Ocean pool is updating a parameter on their mining node that doesn't require a consensus agreement to change. The change only affects their node and the blocks they mine. It is not forcing a change on other miners.
Samurai uses that same function to verify fees are being paid. However due to their specific use of it, they happen to fall beyond the threshold set by oceans parameter change.
The result is that samurai transactions are potentially being filtered (most likely unintentionally).
I say potentially because ocean has not mined a single block since implementing this change and therefore has had no impact on the bitcoin blockchain, yet. Exactly zero coinjoins have been censored or whatever you want to call it.
My opinion: this is sort of like legit emails getting filtered into your spam folder
OP_RETURN - block size #bitcoin
