The Illiberal Liberalism of Charter Cities

I think it's an idea worth trying, though I share some of Tyler Cowen's concerns.
Romer says good rules make countries rich. But countries with bad rules, because they have bad rules, often have no clear path to good rules. Romer says what countries with bad rules need are new rules for generating better rules. His proposed solution is to give up on both the purely endogenous development of better rules and the attempts of the global Lords of Poverty to bribe rulers into imposing better rules. Instead, Romer wants to try to get rulers of countries with bad rules to cede to better rulers effective (but limited) political authority over small, largely unoccupied bits of state territory. It strikes me that there's still an obvious problem here. Why won't the bad rules that have impeded endogenous development also impede the adoption of a higher-order rule-reforming rule? I don't really see the loophole that Romer needs to get started. Anyway, the idea is that the rulers of screwed up countries will be so impressed by these zones of high economic performance that they will seek to replicate them in the territory they haven't leased to Canada or Belgium or whomever.
Hong Kong and it's effect on China is Romer's big example. Alex Tabarrok says that Hong Kong's reintegration into China really marked China's integration into Hong Kong. I think this is too hopeful. China remains authoritarian, illiberal, undemocratic and not at all enamored of the distinctively English spirit of laissez faire behind the Hong Kong experiment. I think it's interesting that these facts are clearly assets to the Charter Cities project. What the example of Hong Kong communicates is that authoritarian, illiberal, undemocratic regimes need not feel threatened by semi-independent city states with working “liberal” market institutions. It says to rulers that their countries can get rich without granting their subjects real freedom.
What should we make of this message? Should we encourage it? Is Romer trying to encourage it? Does Romer believe it? Or does he believe that high growth rates sooner or later lead to broader liberalization. Maybe it's OK to let this cat out of the bag as long as the pace of liberalization is slow enough that current illiberal rulers are never really threatened by the liberalizing externalities of charter cities. But is there any way to credibly make this assurance?
I'm convinced that it would probably be better for both the liberty and welfare of the Burmese people, for example, if the junta tried to go the Singapore/China market authoritarianism route rather than hold free elections and establish a democratic government. I'm not happy with this conclusion. Unlike many of my libertarian friends, I do not think democracy is incidental to liberty. But suppose it turns out that democracy is incidental to economic growth — that it is correlated with but unnecessary to growth. Suppose further that illiberal rulers will welcome isolated experiments in the institutions of growth as long as they don't come bundled with democratic institutions. If economic liberalization eventually has liberalizing political spillovers, promoting democracy directly could turn out to be self-defeating. Could it turn out that liberal democrats do the most for liberal democracy by promoting market authoritarianism? Would this make Naomi Klein right or wrong?