Blockchains are a great idea but some problems are better solved by the brick-and-mortar legal system.
There still exists a legal system in which the NFT could potentially recovered. Yes it's harder with easy cross border theft, but that just means we need better global treaties.
If your car is stolen, there's a potential for it to be physically found and returned.
If someone somehow finds the private key to steal an NFT, they could steal the NFT and then delete the private key for the new address that owns it. At that point, treaties don't matter. The ownership exchange is permanent until someone can derive the private key from the public key, which if someone did, then ALL NFTs could be stolen.
There's also an easy fix for this. Make the NFT include a smart contract that requires periodic "checkins" that require the private key, and if the checkin fails, ownership reverts the previous owner or another address specified in the smart contract.
Or attach a smart contract that requires registration with a third centralized party.
NFT technology can interplay with existing legal frameworks and protections with the right programming.
The global DNS namespace is heavily controlled. I very much doubt that these funny namespaces will be anything other than a fad used in niche applications. Metamask supports .eth, fine, but you wont be pointing at your https server with it if you expect grandma to resolve it.
We have seen it before, many times. NewNet isnt a thing this decade. You cant invent arbitrary TLDs and expect the internet to use em.
IMHO they will fail. You wont see decentralized name registration in the global DNS namespace. Too much abuse potential, too little flexibility with errors, trademarks, etc
I agree with you there. The key problem is maintaining ownership of a valuable domain, not freeing domain ownership from ICANN or automating dispute mediation.
Putting domain ownership on a blockchain just makes it almost impossible to maintain ownership of the domain if someone steals your "crown jewel" certificates.