Physically it is a comparatively a cheap system to run without the high cost of a proof of work system.
Not true. We have to pay for all the bloated government, VISA fees, and debt that central banking sustains.
Proof-of-work is much less expensive.
Correct but that is only our present condition. In theory without all the bloated government, VISA fees, and debt (stripped of all the fat) the actual technical cost to run a lean centralized fiat currency is (theoretically) cheaper then a decentralized cryptocurrency. I will concede that this condition is nothing like the world we live in today. The point I am making here is that in theory fiat may become cost competitive again in the (probably distant) future once the waste is stripped from the system. I will clarify the OP in this area.
Emphatically disagree.
The Communists always argue ideologically that it is more efficient to not have competition. However reality has proven their ideology to be more inefficient than decentralized competition.
Have you compared the uptime and reliability of a Microsoft server versus a Linux server. The
Bazaar (decentralized, open source) model trumps the Cathedral (top-down, closed source) model for large scale systems. I wrote about this having the
only known positive scaling law of software engineering. Only vertical markets such as finely tuned graphical software has
so far resisted the Bazaar model.
This appears to be our last remaining area of disagreement.
I am not arguing that fiat currency will in a utopian (communist like future) become dominate once the inefficiency is ironed out.
My argument is that in the future centralized fiat and decentralized anonymous cryptocurrencies are likely to both exist in competitive equilibrium.
The existence of this competition will drive progress forward in both currency types. Centralized fiat systems perhaps released by a world government or trusted corporation will try to copy the best ideas from the decentralized systems while attempting to leverage economies of scale (lack of proof of work, single server). These advantages will be offset by the the inefficiencies that occur with the cathedral model, the tendencies of all centralized systems to fail, and the power vacuum. Fiat currencies that successfully exploit economies of scale while restraining their natural tendency to abuse their vested authority may survive. Fiat currencies that do not will cease to exist as people vote with their feet and move to decentralized options.
Similarly, competition will drive progress forward for cryptocurrencies. As you pointed out many of the current advantages of fiat can eventually be replicated in cryptocurrencies through innovation. Competition with fiat will drive the transaction time for cryptocurrency down over time. Decentralized cryptocurrencies will undoubtably be safer stores of value without the risk of trusting a central issuer. This safety comes at the expense of directing monetary inflation to pay for proof of work.
I can not think of any advantage for a top-down digital ("electronic" is not correct term, as digits can be written on paper wallets) fiat. Even it would enable the criminals to take over everything.
The only advantage of a top-down digital currency is that it would allow some monetary inflation to be used for purposes other than proof of work. For example, monetary inflation could be used to pay for the necessary constraints on the dynamic system (aka pay for the cost of addressing market failures until a free market solution can be found). This efficiency can only be achieved, however, in the setting of a robust decentralized alternative or the fiat system will collapse for the reasons outlined in the
Economic Devastation thread.