In fact, Intel and AMD didn't profit a lot in this.  If there's one company that even has production problems because of PoW, it is Nvidia, and not for bitcoin, but for ethereum and a few others !  Also, I don't think it is Big Oil that profits, it are the Chinese charcoal mines.  Mining is to be seen quite more literally  

What is startling in this story, is that one almost has the impression that Satoshi's intellect is waning.  His premonitory views in November 2008 were much more to the point than his "incomprehension" in 2010.  Is this because Satoshi was bright, but easily influenced by a dumb ass big mouth that talked him into nonsense ?  Is it because the Satoshi of 2008 was not the Satoshi of 2010 ?  Was Satoshi ill ?  Was he consuming drugs ? Did he get hit on the head ? Or is this the obvious impression that you get from someone trying to trick you into something (a set of lies always look less smart and coherent than a truly brilliant idea) ?

You can't be smart to a point of foreseeing, without one single working system in nature, how thing would evolve in 2008, and be obtuse to the point of failing to see the obvious when it is even spelled out for you by other people in 2010, unless you are someone else, or you were hit on the head, are ill, or are lying.

This is what occurred to me recently ; until now, I simply thought: "well, a big mistake in 2010, but everyone makes mistakes, Satoshi was a bright guy but not a genius".  But now, as I said, the pieces just don't fit together.

What is remarkable, even though nobody ever said it, is that bitcoins' system itself has an indicator of the necessity of the block size: the difficulty !  The difficulty measures the market value, the size of the user community, the technological advances, and the investment mining nodes make in PoW (and the rewards in value they get).    It is quite obvious that things like storage costs and network burden, and their economic cost, are going to be proportional to this.  So it would have been rather obvious to scale block size with difficulty.
The bigger and more valuable the network gets, the more NEED there will be for transaction room, and the more MEANS there are to afford storage and network capacity.  It would also solve the "end problem" of no coin emission, to a point: if fees are too low, and finance PoW, difficulty falls, the blocks shrink to a point where the fee market kicks in, until an equilibrium is reached where the block size allows for a reasonable fee, paying for a reasonable amount of PoW, giving rise to a reasonable difficulty, as a function of demand for transactions, and market price of BTC.