Here is one of the main talking points I'd like to discuss:
Rockefeller/51% attack: Rockefeller was one of the wealthiest Americans of all time. He did this by running a monopoly (Standard Oil Company, Inc.). To create the monopoly, all he had to do was operate at a loss for short periods of time. Although his assets and capital allowed him to enter periods of negative income, his competitors could not survive the loss of revenue and would leave the market. This allowed Rockefeller to retake control of the market, and bring his prices back into profit range again. Rockefeller didn't have much competition, because anyone entering into the market would know already the losses they would take for a risk they weren't sure to win (they wouldn't). The monopoly consolidated power quickly. And because the company grew so large, it was always guaranteed profits in the long run. It didn't matter how much it hurt short term, when the monopoly was reestablished the promised flow of profits would begin again. Selling at a lower price wasn't a long term strategy it just served to eliminate other businesses. This was how he maintained the monopoly.
What about bitcoin, at the protocol level?
Suppose some entity (a government, a cabal of miners, a worldwide energy conglomerate?) decided to band together and they controlled 51+% of the hash rate. They take extreme profit losses in the short term, so that it denies other miners the revenue from getting their blocks added to the chain. The blockchain thus eventually becomes constructed only by the entity that now has taken control. Meanwhile, no other miners are allowed to enter the market because they cannot bootstrap the appropriate funds to overcome the monopoly. By protocol, the longer chain is always be the preferred one.
Suppose this cabal owned 70% of the mining hash rate. You would be providing 70% of the power of the network, but at the same time consuming 100% of the rewards, hypothetically. Joe Kelly thinks this possibility could start in the ASIC chip world, or perhaps be a state sponsored (clandestine) effort. In fact, the path towards a 51% majority would be profitable along the way, but only once you attacked the network would you have to worry about revenue.
An interesting side note: A monopoly could even exist in secret, whereby the main blockchain censors the blocks of other miners, but in such a fashion that a pattern cannot be easily established. Frustrated by a lack of transparency, these miners would then become suspicious and unreliable when compared to the hidden monopoly.
How does this tie into the idea of bitcoin as a projection of power? Determining the fate of the worlds currency (or however you want to define it, now or in the future) is where everyone wants to be. It's good to be king. Why not mine your way to the top, then decide who to kick out? Kick out any country you don't like… or better yet, just slow or inhibit their capabilities on the network, in an either explicit or obscured way.
I understand the energy requirements, but what about a project that started with the miners first? Or what if an energy company decided to tie in all their excess energy to the network, and then work with a government subsidy? Didn't Adam Beck say something about all the unused, excess energy that exists right now could easily increase the size of the network to possibly bear such a brunt?
BTW: I'm pro-bitcoin, never selling, long term hodler, there is no spoon. This argument seemed to just hit me a little differently because I never thought of the possible profits during an attack, not just afterwards.