I often wonder about this as well. How the fees will be regulated after the bitcoin protocol's block reward is negligible? I think OP has made the assumption that the BTC fee volume stays constant. As far as I am aware, this is not the case. If it was the case, then their would not only be a correlation between BTC price and transaction fee (which insinuates some variance) but it would actually be a linear function, such that:

transaction fee = BTC fee volume (constant) * BTC price

either that, or I have missed something in the conversation.

Having said that, what are the driving forces for BTC fee volume?

I imagine that there is the amount of transactions that are being posted to the mempool. Therefore people would compete with higher BTC fee volumes to have their transactions posted first. but then what if there was a dip in transactions posted to the mempool and the fees drop?

In the future when the block reward is infinitesimal and the fees drop, then the miners are getting less money. Which would cause them to lower the hashing power (electricity spend) on the network. Would that lower the security of the bitcoin protocol?

So in the future, would it be that the fees should be regulated by the amount of money that the community believes should be spent on securing the network. Given that there is about 3000 transactions. It would be how much we want to give the miners in order to secure the network such that our storage of value is protected from 51% attacks etc.

What would that figure need to be, to ensure a secure creation of new block? if we know that, then we can divide that by the amount of transactions in a block and that would be the fee. Right?