These are gorgeous and the idea of physical tokens has a rich history in Bitcoin. Some users love them but others not so much. There are two problems to solve:

These are gorgeous and the idea of physical tokens has a rich history in Bitcoin. Some users love them but others not so much. There are two problems to solve: (1) securing the cryptographic material (aka "private key(s)") stored on the physical token; and (2) securing the cryptographic material not stored on the physical token. I can't speak to (1), but (2) has throughout Bitcoin's history been a very hard problem to solve. Even if the token is the only place that the cryptographic material is stored in all the universe, proving that fact is practically impossible. Unfortunately, the page is light on details and the comments haven't answered al my questions so I'll speculate a bit. It appears that this approach goes in a slightly different direction than others with "multisig." This implies to me that both issuer and user generate private keys. This could be a 1 of 2 (either user OR issuer must sign) or 2 of 2 (both user AND issuer must sign). If 1 of 2, the note issuer can in principle always spend the money at any time. If 2 of 2 the user needs to secure the private keys or risk being locked out forever. Either way, we're right back at the fundamental problem: how to secure those keys. If the funds can be lost through loss of user keys, then this offers little practical advantage over bill-less bitcoin. If the issuer can always sign unilaterally, then users become dependent on a trusted third party, which is exactly the problem Bitcoin was created to solve.