True, I did not take into account the initial acquisition value (which I would estimate at 10.000$ in the example above). I only took into account the depreciation of that hardware (and then added another $ per hour). I think that's not unreasonable and standard procedure in profit/loss calculations. The same trick could be pulled of multiple times with that very same hardware, or the hardware can be used for different purposes (e.g. video rendering) afterwards.
I don't think the attack could be pulled off multiple times. Either the value of bitcoin would plummet, or the network hashrate would increase significantly prevent future such attacks. The attack can't be a recurring source of income.
Then who would be paying for that significantly increased network hashrate?
I do agree there's a risk that bitcoin value would plummet, therefore my premise 'a truly valuable bitcoin, with a value that can be depended on'. So either the example shows such a scheme would be profitable, or that bitcoin value can never be depended on.