1. Regarding the mining problem in your scenario:

There's a globally a great variance on out of pocket expenses on electricity prices ($ 0.00 .. $ 0.40), of which near zero could consist of wind / solar / hydro installations.  Also the environment (tropics vs. nordic) will add to expenses or savings on cooling, not speaking vast differences of management, housing and financing costs. Hardware prices won't matter in this case, since the hardware is already up and running in your scenario.

Next to these marginal costs, there's a large range of power efficiencies of mining gear, together with dynamically power saving settings like undervolting it.

Imho, these two factor mitigate the pace of decreasing hashrate to a great extent and smoothen it out.

Also, 'old' mining companies go in stasis till better times or go bankrupt. Other, better situated and more efficient companies buy up their gear for their operations at an appropriate market value. The zero sum game mining is, will deliver more BTC to the remaining companies, which compensates them for the lower market value of BTC. Lots of negative feedback loops to decreasing computing power in the network.


2. Regarding the plummet of the exchange value against the USD:

If BTC reaches those height, many real economy and financial contracts will also be stated in BTC, so there's also some guaranteed exchange and buffered value.

When block times increase, trading is slowed down, new supply of BTC from miners decreases and most humans are anxious anyway to take any loss. Demand could plummet, but so will the supply to the market.