And yes, this implicitly makes Bitcoin a bad idea, as well as blockchain in general.
For example, on Linux the true HEAD is where Linus says it is. Everyone else is wrong. That's a consensus of everyone working on Linux.
In more distributed projects, HEAD is where everyone agrees HEAD should be.
Human consensus is a valid form of blockchain consensus.
No, Linus is a form of an oracle here. Linus says the head is X therefore it is.
(oracles are fine! we could also just call them "authorities"! having an authority at the head of a project is fine! but it's not really blockchain.)
Not if you want a trustless system. Otherwise, why waste time/energy? A bank can currently perform the "human consensus" part already (ala credit cards and fiat money).
Price goes up > People turn on more miners > Difficulty goes up > Less efficient miners become unprofitable > People turn off more miners > Difficulty goes down.
It's a self-correcting dynamic purely due to economics, not linked to usage.
Do you not agree that, loosely, there is a correlation between transaction throughput/ popularity/ real world usage, and the amount of people being happy to invest time, effort, and money into mining?
Is this market failure? I don't know... Bitcoin still works despite it.
I recently had a discussion with a friend about the viability of a distributed/ blockchain alternative to YouTube. Extrapolating BTC inefficiency to a distributed CDN processing TBs and PBs of traffic...
There's about 6 orders of magnitude difference in transaction cost. Mastercard/Visa would use ~2W.h per transaction, Bitcoin about 1MW.h.