> Cryptocurrency allows reliable recordkeeping of transactions without a bribe-able human in the loop. Therefore, it allows default transparency and reduces corruption. Every government gains a benefit from allowing that.
Why would you think a cryptocurrency allows "reliable recordkeeping"? We can put anything we like in a blockchain, including fraudulent records, false provenance chains dating back however long we like, etc. There's nothing to prevent us making multiple blockchains, asserting various conflicting histories, so we can later point to whichever one we find most convenient. It's so easy to make blockchains that there have even been Web sites which will do it for us, e.g. https://web.archive.org/web/20140228160536/http://coingen.io
Perhaps that's 'cheating', and we should only talk about 'established' cryptocurrencies like Bitcoin and Ethereum? It's certainly much harder to spoof or duplicate their blockchains (that's their key feature, after all!); yet they're still awful for recordkeeping, since they're publically-writable and general-purpose. For example, any time a record needs to be published to the Bitcoin blockchain, there's nothing to stop us publishing multiple records, with all sorts of inconsistent contents; we can do this from multiple pseudonymous wallets, and we can even build up meta-merkle-trees between them over time, to weave a bunch of complicated, entirely-fictional narratives. Again, we can point out whichever of these we find most convenient at a later date. If these records are encrypted, then all of the 'backup' histories we created are indistinguishable from day-to-day bitcoin traffic. If the records aren't encrypted, we can blame their existence on impersonators (we can always prove the 'proper' records came from us; and nobody else can prove we also made the 'backups').
Perhaps that's 'cheating', and we need to tie all of the records to a single wallet, to avoid such 'retroactive construction'? That would require some sort of trust/consensus/distribution mechanism so all parties can verify that they'll be using the same wallet ID. Blockchains don't help us with that (the same problem would arise with that blockchain, and so on); but thankfully it's pretty easy to solve in a trustless way: we make lots of copies of the ID, spread out in multiple locations, including public archives.
However, if multiple parties are able to agree on a wallet ID, then they don't need a blockchain in the first place! The records can just be signed and appended to a merkle database, with no need for proof-of-anything. The consensus mechanism is used to agree on a public key, database location and initial hash. After that, each party is free to check at any time whether the database has been tampered with (i.e. whether their hash is still an ancestor of the latest record). They can also replace their remembered hash with the latest one at any point.
People are doing this all over the world every day. For example, that's exactly how git works (+ commit signing).
> As for Bitcoin specifically, in multi-country deals it offers a neutral currency that neither party has more control over. Among other things this allows loans to be brokered without a risk of either party manipulating the supply of currency in which the loan must be repaid. It's a huge benefit to most countries to have a tool like that available.
Lol; there's no need to manipulate the supply if we can manipulate the price. If a multi-country loan required a country to pay a certain number of bitcoins, that country would be incentivised to affect bitcoin's price by enacting laws, spreading propaganda, and generally trying to spook speculators. It's a laughably bad use-case.