It'd be a exercise in futility to implement/create something based on modern cryptography and not be a fan of djb

We did receive feedback on the concept in general and draft in particular. We reworked the crypto significantly from the initial draft, both the key blinding as well as the symmetric encryption. Notably, the crypto is now more agile (you can add new zone types if applicably cryptographic primitives emerge that allow for this particular key bliding)