The hardware to perform PoW already exists. All someone needs to do for a 51% attack is buy up the old mining hardware that is being sold for so cheap.
Every "bust" period for BTC comes with the risk of a 51% attack / centralization.
If the difficulty falls by 90% because 90% of miners were shut off, you only need to buy 5.6% of those thrown-away miners to cause a 51% attack on BTC.
It doesn't matter if they buy it or not. If they're waiting for profitability, they won't turn those machines on. Or do you think people will spend $100 on electricity to mine $10 of BTC just to keep the security of the token high?
A hypothetical 51% attacker isn't doing it for profits, they're doing it for some other reason. Whatever that reason is, the 51% attack gets cheaper-and-cheaper each time these "bust" cycles happen.
Let us know when you figure out that "some other reason." The primary defense against 51% attacks is waiting the amount of block confirmations that make the cost of attacking the block your transaction is in more than the possible rewards. If we all have to worry about people randomly burning money to 51% attack blockchains then the whole premise is bunko anyways.
Here's a hint, randomly attacking the network in a way that loses you money eventually removes you from the playing field. The network is stronger than a motivated attacker, in principle.
I don't need to identify a reason. I simply need to point out that BSV, BTG, and ETC have all suffered 51% attacks already. So someone out there is motivated to perform these attacks.
Every "bust" period for BTC comes with the risk of a 51% attack / centralization.
If the difficulty falls by 90% because 90% of miners were shut off, you only need to buy 5.6% of those thrown-away miners to cause a 51% attack on BTC.