Impact of 30 more halvings

I seem to remember every halving being approached with some trepidation as to how the miners would adapt. Historically, price goes up and more hash power gets turned on. Can this happen another 30 times?

In my simple understanding, each halving will double the cost of production. For most commodities, the price will reach equilibrium near the cost or production. Assuming 30 more halvings, assuming a current cost of about $15,000 per BTC (which I think is a low estimate), this puts the bitcoin cost of production at ($15000 x 2^30) or over $16T per BTC in 2140.

This seems kind of absurd. Especially since it assumes the difficulty and hash rate stay constant to what they are today in 2022. Given the trajectory of hash rate, it should be safe to assume the difficulty will be much higher in 2140 than in 2022. This would mean the cost of production is even higher than $16T, and the market price would need to be higher than $16T/BTC to support this. That would put the market cap at 21M x $16T or $336 Quintillion. Again, an absurd number as this is in todays dollars (inflation would make this even higher).

Another possibility is that the hash rate reaches a maximum and then starts to drop off. If the difficulty is too high, miners will have to drop out of the race and reduce the difficulty (and cost) to be under the market price. This would lead to centralization, or at least a reduction in the number of entities controlling the mining hash rate.

I understand that at some point, most of the block rewards will be in transaction fees, but how does this impact the overall economics of cost of production? The fees are currently running around 1-2% of the block reward, though when the network is busy it has spiked to 30-40%. From what I can tell the fees will stay relatively the same in absolute BTC earned, but the coinbase reward gets reduced with the halvings. This seems like it would put a cap on cost of mining a block thats something like 100x what it is today.

So this would put cost of blocks around $1.5M, and since that's for 6.5 BTC today... that would cap it around $9.75M cap to BTC? More realistic but not sure the math holds up...