But we know it's not true - the government does issue fake passports, e.g. for spies, maybe for other purposes too.
> each individual only has one physical passport
Passports can be lost. In that case, the government keeps the database of revoked passports (IIRC in cases I've seen they actually track when the latest one was issued and will reject ones with earlier issue date). However, if you don't have access to that database, I don't think you have any way to distinguish between passports issued to the same person.
At that would be true for pretty much every country I imagine - I don't think any country would not allow their citizens to replace a lost passport.
Another thing to implement as well is to increase the difficulty for a passport to find a new block once it has already found a block. This way even if someone extracts the private key of a passport he will only be able to take a very limited advantage out of it because the difficulty for him would skyrocket, as a consequence the reward of doing this should be lower than the costs of doing the hardware hacking.
But countries are a completely arbitrary distinction. Should Liechtenstein mining 100 blocks raise its difficulty as China mining 100 blocks? If not, how should we weigh each country? By population? By GDP? What's stopping countries from gaming their statistics to get more votes?
How does increasing the mining difficulty affect the time it takes to scan a passport in line at the airport? It seems counter-intuitive to make passport processing take longer for higher-frequency travelers.
It doesn't. The idea is not to change the actual passport, is to change the software that uses the passport to perform these calculations. That software would have to ask the passport to do 10000x operations instead of 100x. But each would take the same time as before.
Yes, both require some level of trust in governments, but that doesn't mean they are the same.
Representatives from each country are voted in by the people, and those representatives vote...
(for now assume all countries have perfect democracies and the UN does away with its veto system)
Overall, it works, but I suspect we wouldn't see the explosion of "Blockchain 2.0" companies if instead they were "UN Governance 2.0" companies...
But this could be interesting as a sybil-resistant voting mechanism. It could be helpful for experiments in quadratic voting, liquid democracy, etc.
As you point out it's not perfect, which is why I used the word "experiment," but it's a step forward at least.
This idea goes against the decentralized, trustless and permissionless nature of cryptocurrencies.
The meat of the idea, as I understand it, is: passports can be easily validated (because they contain a secret key signed by a government's Document Signing Certificate), but they cannot be cloned (because that secret key is in tamper-resistant hardware). So let's prove work by generating preimages of signatures under these keys.
Trusting this purported PoW system requires that you trust the government not to issue fake certificates. Otherwise, they can just use their signing key to generate lots of new certificates and massively parallelize the "proof of work". Moreover, at the point that you trust the government to behave honestly, there's no need for proof of work! Since you're trusting that a passport key is uniquely tied to an individual, you can (for example) just vote---the required assumption already implies that there cannot be any Sybils. (Or use any other mechanism that requires Sybil-free PKI, since that's what the assumption implies.)
(As another practical matter, there does not appear to be any reason to believe that each individual only has one physical passport. Some countries allow people to have multiple passports.)