I have to admit that I have no idea how much work is actually needed to secure the network. My point of view is that the current rate of energy expenditure outweighs whatever benefit Bitcoin does or could provide to society. But if this rate is a transient result of still-significant minting going on, things could definitely look different in the future.
Do you know of any analyses on how much work really has to be continuously expended in order for Bitcoin to remain reasonably secure at a given market capitalization?
Each nation would love to be able to manipulate the supply itself—why not, if people will let you get away with it?—but the fact that other nations can't do the same could be seen as a feature.
If Bitcoin does eventually become a common instrument of trade at this level it will fill the same niche currently occupied by gold and other precious metals.
The only real problem with that is that with a small hash rate, bitcoin can be attacked more easily.
If bitcoin is the monetary backbone for many nations, they will subsidize miners to maintain the balance of power. That is the actual scenario that I'm optimistically predicting.
If bitcoin isn't the monetary backbone for many nations, by then, then it's probably a failure, and should probably be allowed to die.
It's also very possible that transactions fees alone actually will be sufficient to support a high enough amount of hash power to secure the network.