Another misleading scumbag spews lies and have guts to lie and claim it’s verifiable on the blockchain.

Unmixed change is not part of the coinjoin period.

In fact, it's because of this lack of tx0 that you get deterministic links in mixes. The lack of segregation of unmixed change outputs -rather as part of the mix - adds to this demixing risk. That is why zerolink protocol has tx0 to segregate unmixed change from coinjoin tx.

Tx0 is not linked with postmix derivation path of Samourai wallet. That is why there is a premix derivation after tx0 happens.

You seem to have no idea why tx0 is there for.

There is no way unmixed change can end up in the coinjoin or postmix side of the wallet user must physically send that change directly to postmix derivation path.

In the Sybil attack case you mention which would be costly but then we can assume the attackers would use a special SCODE that gives them a 100% discount but then it would cost them miner fees just to perform this attack which should not be that expensive. If this is likely the case then no CJ implementation is safe since Sybil attacks like this are possible but since Whirlpool coordinator is blinded and the more you mix the higher the probability increases and the more self hosted dojo mixers join to increase your anon set would make the probability a lot harder hence why dojo server was released so to decrease the possibility of SW from hosting everyone's *pub. I have honestly not been able by myself or with the help of others who can navigate inputs/outputs to check on backwards history after even performing one mix so until I can see someone do this I will continue to use whirlpool