The velocity number in the formula is constrained to the BTC used within transactions and not the BTC used as store of value, which would heavily decrease the number. I'd guess it would be closer to 10-25 range. Additionally, any coins lost would permanently skew velocity down. Parsing the blockchain also presents a lot of problems because not all transactions, especially large ones that skew network statistics, have any value. Bitstamp might modify their wallet management systems and redistribute 500,000 BTC 100 times. I avoided bitcoin days destroyed for this reason.
I'll definitely check out Metcalfe Value. Thanks for the feedback!
Fair response, as your method attempts to separate "active" from "inactive" bitcoins. I think you should make your tables clear, however, and call it "Velocity of Active Bitcoins" as opposed to just "Velocity."
I'd still like to see historical data on gold velocity. I expect the same phenomenon happened in pre-fiat times: the rich dukes had vaults of "inactive gold" while middle-class merchants traded "active" gold at a much higher velocity. It would be
very interesting to know the ratio of inactive/active gold over history and the corresponding velocity.
EDIT: I guess there is also a straightforward mapping between "velocity" and "velocity of active bitcoins," that just depends on velocity and the percentage of bitcoins that remain inactive, correct?