A miner is running a business and has made a significant investment. Why would a businessman kill his own business by attacking the very thing that is making them profit?
Yes, this is already taken into account in the estimated price of the attack.
I'm not talking about the costs of the attack, I'm questionings its very viability.
Also, again,
the paper is assuming an attack duration of 1 hour.That's $6-20 billion thrown out the window for... what? Showing that it
could be done? The thing is, the 6 confirmations that the paper is using as a basis for calculating the attack cost are pretty much arbitrary. In case of an imminent or ongoing 51% attack, exchanges and merchants can simply increase the required confirmation count, increasing the attack's cost signficantly, with very little effort.
So basically you'd spend tens of billions to stall the network for a few hours. Basically what Ordinals achieved by accident, but to a much lesser extend.