Thanks @Musyimi97 … that’s a helpful position on subject identifiers, and I agree: alignable by convention, with the lease layer keeping its own identifiers.
The cross-registry ambiguity point is well taken; we’re looking at an optional extension exposing (chainId, registry, anchorId) rather than overloading a single reference field.
I’ve posted the concrete mapping table you asked about here:
ERC-8348: Financial Lease - ERCs - Fellowship of Ethereum Magicians
The ERC-8330 portfolio case turned out to be the most interesting one … for inflation-indexed leases, valuation staleness maps directly onto the index oracle’s asOf timestamp, which is a stronger binding than I expected when we started. Happy to go deeper on that one if it’s useful for the family.