I'm not comparing anything I'm just showing the trend and those prices are reported by the time of halving. Feel free to use any average price you like to see the same exact trend.
Indeed, you're picking dates between different halvings and compare the miner's income based on the price of each day. The problem with those examples is that the price was heavily fluctuating. Each halving affects the supply decreasingly compared with the previous one. If you notice it, the first halving made the highest increase, from $10 to $1300.
If the price keeps moving that way in each halving, yes, the security remains the same (or increases). But, the price won't keep moving likewise and I believe we should take it for granted that after a bunch of halvings in the future, the price will start stabilizing. The more the market cap becomes, the more the stabilization.
So, I repeat. If the price stops being so volatile, the computational power offered to the network will slowly decrease. This will happen sooner or later whether we increase the block size or not.
What we can't predict is what will things be like after 100 years, saying anything about it is meaningless in my opinion.
If you
can predict what will happen in the far future based on the present's constants and you prefer not to, then it's inability to face the truth rather than meaninglessness.
It has never really been the case at all. Each of the halving are so far spaced apart in time, there is always a more than sufficient time to ensure that the compensation is sufficient such that the difficulty drop is less than proportionate.
In the first halving, the price did a 130x. In the second, a 10x and in this one, a 7x 'til now. That's why it has never been the case.