Thanks for the article, it was quite interesting as I don't have a US background. I was a bit surprised that the author thought that people might assume that a system of explicit rules, judgements and binding agreements could be anything other than political in practice.
We use language to create these legal instruments, and interpret them as language users in a community, each of us with our own interests. Politics is what we call this interpretation, negotiation and transformation process. But the fact that there is a political process (whether you agree about the current scope of these state-sanctioned instruments, or whether you'd prefer that civil society manintains them instead) is a classic marker of a rule of law society: totalitarianism, as the extremis case you describe, has very little of this political process and for very good reasons - it's not necessary in the broad.
I was not advocating anything, merely describing what often goes undescribed. Further, I wasn't arguing from where the legitimacy emerges, only that it exists, and that that extant legitimacy, coupled with the extant power and willingness to engage in otherwise-criminal acts is what makes a government a government. Take away one or the other and we wouldn't call it a government.
>I quite like rule-of-law states
Here ya go, a long but rather good piece written by a former professor of mine:
http://faculty.msb.edu/hasnasj/GTWebSite/MythWeb.htm