The relationship between increase of reward and increase of difficulty is not necessarily as you say. It's somewhat similar to "what was first, the chicken or the egg?".
Can the difficulty increase once it has his miminum profitability without an increase in price? No, it can't.
We had tens of those evens in the past, mienrs staing on the shelves as there was no incentive to add more hashrate since the profits were marginal.
Have we seen difficulty spikes long after the price has stopped growing? Yes we have again seen it numerous times, look now, the overall difficulty despite those flops is still up 6% and Bitcoin is not at all near the ATH, the difficulty keeps rising and it will keep rising even if we drop to 30 the break-even point is well below the current price for the number of miners we have on right now.
There is no chicken and egg, the difficulty follows the reward, sometimes it might now keep up p with it but never has the difficulty dictated the price.
With the current semiconductor shortage the situation is unnatural an indeed the difficulty doesn't rise as much as expected, but, on the other hand, isn't this yet another sign that the overall rule doesn't always work as expected?
It's a freak situation, simple as that, it can't be a perpetual state of affairs and its effects will soon be negated.
Bitcoin will consume as much energy as the block reward will allow it to, if it drops to 100$ the consumption will drop 100 times too
This is non-sense. Why would it drop 100 times if it was evaluated with $100?
Because miners would only make $90k instead of $45m.
How many miners can you feed with $90k even at 1cent per kwh? Oh, let me answer that for you, around 100k S19pro, so just 11 Exahash.
1cent/kwh , no ROI no cooling, no maintenance, no profit.