Bitcoin-like blockchains do not scale like conventional "distributed systems" are expected to. The Bitcoin network handles 7 transactions a second if there are two nodes or if there are two billion. You get more durability and possibly more trust if there are a large number of diverse nodes, but you don't get more capacity.
The scenario where a few traders who meet around a buttonwood tree (see https://en.wikipedia.org/wiki/Buttonwood_Agreement) share a permissioned blockchain makes more sense. The world might need 10 copies of a ledger, but does not need millions.
I think also people have unrealistic political ideas around blockchains. When I hear "fiat currency" I tune out; bitcoin is not the new gold. People wanted gold 4000 years ago, they will want gold 4000 years in the future, in all fictional worlds except for one (https://en.wikipedia.org/wiki/Element_79_(anthology)) people want gold. People won't want Bitcoins 4000 years from now.
Also people have unrealistic expectations about privacy. Sure, governments cannot control admission to the blockchain, forcing you to give over a taxpayer id, but your whole transaction record is visible to everyone. Thus your local government can see you ate at Subway and put the pinch on them and everyone you transacted with. So can the mafia, some other government, newspaper reporters, extortionists, Scientologists, etc.
Ethereum is kinda cool because people can bake applications into it, but there is no real security story for the applications written on it. We see hacks again and again and will continue to see hacks. The escrow model is good for implementing a sports book or equity exchange, but you could not implement loans, swaps, options, etc. on it.
I liked my answer and some other people's answers there.
Edit: something apparently not mentioned in the other thread is the way current cryptocurrencies consume enormous amounts of power.
While this power is used for a useful purpose in some sense, the economics of Bitcoin hardware in particular haven't produced the kind of decentralization of mining that the original paper seems to foresee, yet there has been an incentive to use tons of energy. So, it would also be great to see cryptocurrencies find a way to do proof-of-work that achieves greater decentralization with significantly less energy cost. (There is of course active research on this and I wouldn't regard it as impossible, but many people see the energy issue as an extremely disturbing source of waste.)
Bitcoin Energy Consumption Index was posted just recently on HN. https://digiconomist.net/bitcoin-energy-consumption
"Money is a religion. Bitcoin offends some people's gods."