Great article, thanks.
And this ^ is exactly the right point. But actually, I would say we do know something about the fee market based on wider trends in the payments industry at large which is on a very clear trajectory towards zero transaction fees.
The Visa/MasterCard monopoly days are over and bank-to-bank transfer protocols with no fees already exist today: it’s only a matter of time before these become a) ubiquitous, and b) the default rails for settling the vast majority of transactions in developed economies.
In other words, we know that zero-fee payment rails are how the vast majority of transactions are going to be settled in the near future, meaning they will therefore not be part of the user-side demand, as you put it, for Bitcoin transactions/block space.
No person or business is going to voluntarily incur an unnecessary cost per transaction if they don’t absolutely need to. And they don’t.
The idea that you can amortize insanely high base layer fees over tens/hundreds of thousands of layer 2 transactions is therefore a red herring: Lightning Network transactions will always necessarily be more expensive than zero fee transactions because they must cover the security cost of the base blockchain.
This is Bitcoin’s intractable problem — it is fragile to the emergence of zero fee settlement rails, and we can already see these rails emerging.
The only minor nitpick I have is with this:
Only a hash rate that rises faster than hardware improvement truly implies increasing security.
There are other concerns within the mining game which would make this “increasing security” illusory. E.g. if most of the hashing power is in too few hands, or can be seized by governments, or the supply chains (energy production) are disrupted or manipulated, or there just isn’t enough proof-of-work expenditure (measured in $) happening overall, then the network is still far too vulnerable, irrespective of hash rate vs. hardware improvement.
I wrote about this topic in depth here, you might find it interesting! :)
