We have evidence in this thread of users not wanting to accept Segwit transactions. That's users not miners, that's a lack of fungibility, and that's an issue.
I didn't know that. If you gave me a legacy BTC address (if we had some sort of deal or transaction), I'd send you some BTC that came from a native segwit address. You'd still get your BTC, and you'll likely get it confirmed in a block faster too. Why would you not accept that? Would you return it?
Its a crazy idea, but it is true. There are some Bitcoiners out there that will not accept a Bitcoin TX if there is any segwit TXs in its lineage. Take this with a grain of salt, as my understanding of this attack is very basic and I have no clue how viable it is. The idea behind not accepting Segwit TXs is because there could be a potential 51% collusion attack in the future, where the miners would roll back the Segwit fork and accept all the past Segwit txs as anyone can spend. They would then continue mining the original protocol satoshi designed. So theoretically, there would be two chains. The legacy BTC chain, and the rolled back Segwit chain with all the stolen coins. Any "Legacy" BTC that has no Segwit TXs in its lineage have no risk of being stolen in this attack.
Again, like I said, I have a very rudimentary understanding of this attack. I don't really know the specifics or its viability. This is something that Anonymint has been talking about on this forum recently under his alt account "anunymint".