The Golden Arches Theory (even before it was disproven by the Russia-Georgia war of 2008) -- and similar theories like the "democracies don't fight wars against each other" theory -- have always been silly things that take rely on people not understanding math. You've got a feature that (at the time the theory is articulated) is historically fairly recent and that, when you take the number of pairs of countries that have been in wars that have occurred in any given time frame and the total number of pairs of country that have existed in the same time frame, and the total number of pairs of country that share the feature in question, where the expected value of number of wars between countries sharing the trait that is supposed to protect against war would be closer to zero than one if wars were randomly distributed and the feature had no effect, and then the theory uses the (utterly unsurprising) fact that the actual number of wars between countries sharing the trait is zero as the whole basis for an argument that sharing the trait prevents war.
If you read the Wikipedia page about the book you'll see Russia-Georgia listed as 1 of 5 such counter-examples, going back to the US invasion of Panama in 1989.
You'll also see his responses.
But I'm not trying to justify either which way. My point is that there are certain concepts in geopolitical dialog that are used as short-hand to express a larger concept. Expressions like "McDonald's cannot flourish without McDonnell Douglas" are to outsiders as meaningless as "Darmok and Jalad at Tanagra" or "information wants to be free".
In other words "[Schmidt] struggled to verbalize many of [his politics], often shoehorning geopolitical subtleties into Silicon Valley marketese or the ossified State Department micro-language of his companions" can be turned around - Assage uses a different language than you or I, though he doesn't struggle to verbalize his politics.
