A more sensible way to measure market dominance of an electronic payment system:

LESS USEFUL: "value" of the system accounting units priced in some fiat currencies

MORE USEFUL: fraction of global commerce transacted in the system, e.g.

• percentage of the number of transactions over number of all global transactions in same timeframe

• percentage of value transacted in some timeframe compared to the other common payment & settlement methods (maybe bring everything into a common unit of account e.g. via gold price)

There would be difficulty working these out. Statistics and surveying would need to be used.

But let's be honest - in today's world, the results would not (yet) be flattering for crypto "currency".

And it might surprise that coins ranked highly according to the current dollar price based market cap metric may rank much lower if looked at through the above glasses.

Now, if A and B are comparably good money, and legal to use in commerce, a merchant should not care much whether they get paid in A or B.

All things being equal, this should help a free market select against the worse forms of money and migrate to the better ones.

Of course, things are not equal - there is the famous network effect of money that factors extremely heavily.

But that - the actual network effect - is much better captured through evaluating the percentage of actual commercial transactions done with some money, rather than using the manipulable statistic of its supply capitalization via heavily inflated fiat.