Well, there's the other tragedy of the commons problem to contend with there, miners see almost none of that revenue from L2s for mining, and no revenue from securing hodlers' wealth for them.
Given that it can be expensive to use Bitcoin L1 and/or spending will be disincentivized because of the hard cap, it becomes more likely that an L2 will be used. However, L2 solutions are typically not self-custody and are subpar custody in comparison to an L1. L2 custody may become an unacceptable risk in the future.
Let's assume that self-custody gets simple enough for the common man to be capable of doing it on their own. In that environment, it would be simple to use the L1 of any blockchain.
Any L1 may become the L1 used for inflationary spending. All major L1s fail to preserve purchasing power like Bitcoin, and privacy like Monero. As chain surveillance becomes trivial, this feature demonstrates the invaluable: cheap cross-chain privacy. Atomic swaps could be implemented, such that it is the main method of transmission of value between XMR and any chain (but let's focus on BTC).
BTC may be used for deflationary savings, with spending overall being reduced, but needs and desires still exist in the physical (3D) world. A refactoring of what those needs and desires are becomes significant. However, people will still want to spend their 2D savings at some point for luxuries in 3D space. It is possible that Bitcoin becomes so valuable that all savings are immediately transferred into BTC L1 to hold. Tx fees can remain low in sats/vb terms, but increasingly more valuable in infinity/btc terms. A few Txs or even one Tx per block may be valuable enough to sustain the network in the future. Assuming humans continue to trade, there should be some Tx occuring to convert other L1s into Bitcoin L1.
This scenario presupposes that L2s and wrapping assets cross-chain becomes infeasible due to 3rd party risk.
