It suits the state to run the land registry: (1) a state is a barrier around land, which can be subdivided; (2) the state has incentive to do this; (3) the state can use law and courts to enforce it.
There are other problems that need registries where there is no party with inventive and means to act as the state does above. I think the physical-art and wine examples fit that.
The key quality of blockchain is that it allows groups to form consensus. Even though they compete on some matters, all stakeholders want the platform to succeed. If it doesn't, their investment is worthless.
Hence, Blockchain can act as a synthetic state for enforcing standards and governance of registries. This may go as far as having elections, rulings. The registry data itself could live off-chain. [Sure, there are details to be sorted out here. Maybe the blockchain entity would control all APIs, to allow it to enforce standards against the registries. Those APIs may not necessarily be on the blockchain.]
Due to incentives, I don't think the trust model is the same as a standalone registry. One of the many challenges of dealing with data providers at the moment is they can routinely have low data quality and still get paid. Whereas a state would intervene if a land registry fell below standard.
"In your other answers, you seem to take it for granted that using a blockchain means everything is interoperable"
This is a real problem. I think the foundation problem to solve is how to get somebody to fill a role equivalent to the state in the land registry example. Once you have that, you have a path to solving issues such as this.
Thanks for poking at this.