Imagine you are waiting for a flight with 100 seats. The seats are auctioned with the 100 highest bids buying a seat. The auction opens 10 mins before closing, and previously unsuccessful bids remain active in the next auction.

The number of new people showing up per flight ranges from say 10 to 300. You are sitting in the waiting area and can see how many people are already waiting at the gate. If you see that there are regularly spare spaces (more seats than people waiting) you pay the minimum possible amount that the desk will process (1sat/byte) regardless of how urgently you need to get the flight.

Once there are consistently more than 100 people waiting for each flight the fee you are willing to pay jumps from basically nothing (1sat/byte) to whatever you value the flight at. You now actually consider how much you are willing to pay for the first time, it now matters whether you are you going on a long trip with no plans Vs visiting an ill relative.

For almost every bitcoin transaction the value of making the bitcoin transaction is much greater than the current fee rate.

Fees (in bitcoin) are going to skyrocket once blocks are consistently full. After some block there will never be “empty” capacity in blocks again.

How long before this happens?

This is obviously unknowable - but I’d be interested to hear some predictions based off modelling the growth in the number of bitcoin users, especially with the upcoming halving.

How can number of users per byte be increased?

Batched Transactions, Vertical Scaling (Payment Channels / Lightning), Horizontal Scaling (Opendime), New proposals (Advancements beyond the current taproot/schnorr BIP’s to increase block density).

Are you prepared to utilize these methods?

Are your utxo’s scattered or organised, do you have some payment channels, do you have a range of sizes of utxo (rather than a single large utxo or many very small utxo’s).