It's an insightful perspective and, along with PG's stance on the essence of economic inequality per se, a welcome antidote to the populist outrage machine, which does indeed the frame issue largely in zero-sum terms.
However, there are some counterpoints that don't receive adequate representation in this account, in my opinion:
One is the size and scale to which rent-seeking behaviour dominate the American economy. PG does acknowledge here and elsewhere that rent-seeking behaviour accounts for the wealth of many, but dismisses it relatively quickly as a seemingly self-evident byproduct of the expected variance in a society that permits economic opportunity. I think the situation is a lot worse than that; the amount of such parasitism, in the form of regulatory capture, lobbyist influence, outright Gilded Age-style purchase of legislation, revolving-door career paths, etc. account for an extremely significant percentage of US economic output and the unequal concentration of wealth. Consider for example how our healthcare system works (even post-ACA), Big Pharma, the military-industrial complex, intellectual property law and software patents, etc. A great deal of our government is for sale, and the sole purpose of a lot of our legislative projects is to route money into private hands, with the support of the government's monopoly on force, while socialising risks and losses onto the taxpayer. In my specialisation of telecom, I have seen this at work with the hundreds of billions in effective subsidies given to the AT&Ts and Verizons of the world, ostensibly to support the build-out of competitive next-generation broadband infrastructure but in fact to line their own pockets. All in all, the total dimensions of corruption at the top of the economic food chain are in the trillions of dollars, and I feel this insight is not given a fair shake with the same diligence as other aspects of corporate-industrial history of which PG treats.
The second issue relates to the optimal amount of economic inequality we can sustain while maintaining social order and an environment conducive to ongoing innovation. Many revolutions and upheaveals in modern history attest to the fact that when a sufficiently large class of poor and disefranchised people arises and is left to twist in the wind, at some point "radical discontinuities" will occur. Pervasive, festering social ills don't serve the self-interest of the wealthy and the middle class, either; they ultimately impact the security of their lives and their property, requiring them to resort to increasingly drastic measures to keep what's theirs. The market for mass-market and/or consumer products--on which a lot of startup business models depend--is inevitably limited or shrinks when large segments of society see eroding disposable incomes. Instability also negatively impacts the transaction of business by making the outer world less predictable and dependable; lopsided opportunity in savage inequality leads to lopsided and inconsistent educational outcomes and, ultimately, a more heterogenous and troublesome work force.
As other countries around the world periodically assert, there may be a formula for state economic involvement, taxation and social programmes that better maximises more desirable social outcomes, even if it comes at the expense of bridling notional economic opportunity and the smoothing out of some peaks. Is it not true, for example, that much of the significant innovation in computing and networking from which we benefit today came out of Bell Labs, a quasi-governmental institution whose decidedly mid-20th century model of economic existence created the right incentives for long-tail R&D?
The work done in such places, as in the pharmaceutical industry for example, is now fiercely subordinated to narrow, short-horizon commercialisation objectives, and while that may be a better way for some actors to get rich faster, is it really where we want to go? It seems to me one can raise the same kind of objection about the incentives set up by unicorn-seeking VC funding and California startup capitalism.
All this leads me to say that perhaps these issues need to be considered in a more global, integrated, and holistic way, rather than narrowly construed as problems of economics. If we want to evolve toward a better future, we may need to take up more thinking from the "normative" sphere, which economists hate.