EP:
That makes sense. It's interesting you say that about CEO's vs. middle managers, because I would almost make the opposite argument. There is clearly a case to be made that good CEO's are even more valuable than ever before. I agree that this is not necessarily a technological case. That said, empirically, CEO value and pay are if anything negatively linked. The best CEO's, the Warren Buffet's of the world, do not seem to really be doing it for the money. This flies in the face of a lot of armchair neoclassicals, but can clearly still make sense from a utility perspective. However, even if you can create an economic model where it makes sense that the best CEO's are not necessarily motivated by money, I'm not sure if that helps the argument that we need to pay ever CEO a shitload of $$.
I would actually say middle managers, as much as they are often reviled, are capable of adding a lot of value and probably do on average. They are basically managing the largest unit of people that a leader can truly impact in my opinion. Decent management practices probably add a lot of value even if it's not sexy. There was a recent Accenture survey that shows that even basic operational management practices resulted in huge gains in Indian small businesses (although Accenture is obviously biased based on their consulting practice).
I do think mitigating the lottery effect of finance would be beneficial overall. I know I would certainly like to get paid similarly to actually make something, but I work in finance for the intellectual challenge and money. The way I look at industries like law, accounting, and finance, is that they are necessary support industries for the "real" economy. Too small, and that probably means you don't have effective property rights, trust in markets, and are under-investing in productive endeavors. But all else equal (assuming you are hitting these goals), you want these industries to be as small as possible as they are essentially support professions for people that are doing actual stuff.
One way you make things like banking less lucrative are higher capital requirements, and limits on their outside business. It seems like they are having some success in this, but I am not even remotely qualified to know 1) how successful it will ultimately be 2) what measures have large loopholes and 3) what measures might actually reduce credit in such a way as to be economically damaging. The other aspect, in my opinion, is to find a way to reduce the lottery effect of investment management. If we are going to let hedge funds market to everyone (which we should I think, since it's unfair to limit niche strategies that may actually work to the already rich), we should also require all hedge funds to post audited returns in a public database. That's it. I think if we had a fair accounting of what the typical fund does, these things would become a lot less interesting to people in a hurry. The available databases are expensive and inaccessible to the public, and so fraught with bias it is easy to convince people what you are doing is worth 2 & 20.