As long as all crypto purchases require an exchange of fiat for crypto the total amount of fiat in circulation never falls. It doesn't hurt if people transact in non-fiat assets (that may be preferable in many cases given the efficiency gains involved). Sure, the growth of alternatives to fiat may lead monetary authorities to reduce the amount of fiat in circulation, but all that is really necessary to prevent deflation is that prices continue to be denominated in fiat and that fiat continues to be acceptable as legal tender.
Yesterday you were complaining that that if everyone transacted in crypto the money supply would be deflationary; today you are insisting no-one will want to hold crypto assets because of volatility. Yesterday you were comparing crypto to the gold standard; today you are suggesting that all economic activity prior to Bretton Woods was "barter" (as if that is an argument!). And then it is clear you haven't even read my previous comment, which explicitly states that cryptocurrency is non-problematic so long as prices continue to be denominated in a non-deflationary asset -- so of course taxes need to continue to be paid in fiat.
So I'll stick to my earlier observation: if you want to understand how crypto will intersect with the fiat economy you need to go back and read Keynes on the causes of liquidity traps, and then work forward from first principles. You don't see it now because -- like Krugman -- you think you are arguing against the Gold Standard. But the Gold Standard is irrelevant: in practice the only question that matters from a macro perspective is to what extent fiat-denominated activity will shrink as crypto-adoption spreads, and whether the manner in which crypto is adopted translates into wage and price rigidity across the economy as a whole. If your generation of macroeconomists do not or cannot understand this, the next ones will.
Firstly, you are absolutely right that by calling all transactions that are denominated in non-fiat assets ”barter” I have inadvertently and inconsistently folded-in all transactions that occurred in official currencies that occurred prior to the abandonment of the gold standard, and that was not my intention. As such, I wish to amend my statement so that it is understood to apply now, far after the abandonment of the gold standard, and relates mainly to the fact that all official currencies that I can think of are fiat in nature, so that transacting in official currency and in non-fiat are understood to be mutually exclusive options (at this moment in time).
Secondly, I do not wish to give the impression that the inherent volatility of cryptocurrencies makes them unsuitable as investment vehicles: given their (naive) return/volatility profiles, they definitely have a role to play in an investment portfolio. What I am trying to say is that they are suitable for investment and speculation, not as a measure of value and as a unit of account (traditional roles of currencies) because they’re too damned unpredictable.
Hopefully you read this and given those two provisos you are willing to resume the (very engaging) conversation.