The answer to the final question (“… why the governments didn’t invent this sooner,” he says. “I came up with this over a weekend in my spare time, why didn’t they? …”) in this here very interesting piece, is easy: Enrollment Problem Plus Risk Management.
But still, the idea of using Bitcoin crypto style solutions to the ‘international’ passport problem is useful, and might work. In some way. Not this self-certification one. If you’re aware of how long PGP has been around, you should be aware of all the failures of any form of tribal-cred-branching-out IDs.
And, a great many governments may just not have a sufficiently pressing need for a new passport scheme. The risks of the current model, are known and (again: apparently) manageable.