If blocks are mostly empty and miners are not confirming 10 Sat / Byte and under fees because they choose to not accept low fee txs than another miner will come along and include those low fee txs as there is an incentive to do so.

If a fee market doesn't exist than there is no one paying a premium for priority block inclusion and much less is collected in fees. A fee market is precisely what allows lower fees to be included at a later time and higher security for those who want priority.

This isn't hypothetical, already Bcash sometimes has less security with Coinbase reward and their fees than Bitcoin fees alone at times. This is why Bcash has had to move away from the whitepaper and is principally using developer checkpoints and only using PoW at the chaintip due to being so insecure. As we have seen with BSV this doesn't stop reorgs or double spends occuring which will start to happen with Bcash too once their hashrate % drops from very low 3-4% to dangerously low 0.5-1%

https://halshs.archives-ouvertes.fr/halshs-00951358/document



Delaying the fee market simply is a kick the can problem ,which would also delay the efforts to work on more efficient and more private 2nd layer solutions. We need to start the slow and gradual transition to fees paying for security vs Inflation. We always knew this transition had to occur. It has been proven that Bitcoin did not need more blockspace because the market of users and businesses still aren't using all the space from the last upgrade segwit(max 50% are segwit) and soon we will see another upgrade of 17-40% more txs per block onchain with signature aggregation.

Any programmer knows to scale efficiently it is better optimize the code first before throwing more resources at the problem and this is what Bitcoin is doing. Upgrades like segwit, taproot, schnorr sig aggregation are exactly what helps bitcoin be more securely scalable in the future when we choose to upgrade Bitcoin from 4 MB of weight to a higher number.