glad to see that you still active. So far I have learned a lot from your old writings.

Since your are both well educated in consensus and in probability theory I have two concerns:

First concern is the build-in a priori assumption that Bitcoin is a store-of-value at all. Free from hindsight: 1634-1637 a tulip bulb also was a solid store for value, but on the ontological level it lags the thermo-dynamically hard-properties of a piece of gold.

Every extrapolation-based model implies, that the scaleability-trilemma is solved ON-chain... which is a strong assumption. The "coin" other than most people think, is an emergent property of the replicated state machine. The coin is not "zeros and ones" and it is not just the sum of all locally stored "copies". In fact there are no copies of the blockchain or the coin, because redundancy is bottom up and never top down. A part of the system (the single process/node) cant hold the emergent property of the replicated state machine. So coins only sit in the grid. Mutualy shared information does not make a copy. A copy is something with the same properties. I mean this is trivial for you but most people still don´t get it.

Their existence is directly dependent on the miners profitability and the miners profitability as far as I understood is dependent on >>ON<<-chain scaling.

Bitcoins correctness is only guarantied by constant validation of the distributed state. It is not quiescent. Even when there are no/not enough transactions, miners still need to validate. Ergo Bitcoins grid-storage is RAM and not ROM. Coins should not be considered as permanent or as Szabo calls it "secure property titles".

Second point is the mathematical rigor: Markets are fat tail distributed right? Hence convergence to the real mean is not observable in real time. Why you bother with those "predictions" at all? Even Nicks co-integration to PlanB´s prediction is based on a violation of the law of large numbers... I mean your reconstruction is rigorous because its not a prediction and has no strong assumptions but the object of your study, the prediction of MCAfee is just bogus?