I expect that's dominated by "infant mortality", because situations that cause currencies to die aren't fixed by rolling out a new currency. If that's the case, it tells us very little about the safety of well established currencies.

They aren't fixed by rolling out a new currency if they're just repeating the previous process (printing money aka inflation, backed by nothing, etc).

Sure, but the ones rolling out a new currency here are bitcoin, which fixes only some of the issues.

Incidentally "printing money" is not "aka inflation" - it causes inflation if the money supply grows faster than demand for the money but they are different things.

A figure that might actually mean something, if we were able to assume that any given fiat currency was the same as another.

Well they're almost all monopoly currencies (not a reference to the game) that are run by people who can increase the money supply.

So basically we can just plug your 27 year figure into a probability density function to find out how many more years we can expect the USD to be kicking around? How many years does it have left, at varying levels of probability?

The incentive of a monopoly is to underproduce.

The Fed and a government that likes to spend has proved otherwise.

Not at all. If production of dollars was competitive, they would be priced near the cost of production. I'm not saying there are no valid critiques of USD and our monetary system. I'm saying that it is not bad for the reasons monopoly is usually bad.