Being a new conceptualization emanated from global financial crisis, can it be utilized to explain the crises historically and help prevent future ones?
Economic Globalization, Political Democracy and Nation State are the three horns. The three scenarios by choosing two of the three options are: Golden Straitjacket, Global Governance and Bretton-Woods Compromise.
It is well known that Latin American crisis countries tended to close their capital accounts in the aftermath of a crisis, while that is not the case for the Asian crisis countries. But during the crisis and its aftermath a country (in development and not in dev) needs a courageous economic policy to foster recovery and mitigate the effect of the crisis on the poor. If you want a socially feasible policy you need capital control to have an independent monetary policy to make your political choices compatible with a fixed exchange rate. If you abandon the exchange rate you risk an inflationary spiral with an increasing value of public debt (specially external, now in an undervalued currency).
If you don't care social sustainability, you can make different choices
Let me take a different tack. Many historians recoil at that notion that models developed in one historical period can be liberally applied to understanding very different historical situations. Those earlier outcomes, they would say, arose from a particular set of causally complex processes and contingent events. This is a common constructivist point of view. Most economists are positivists. We simplify with abandon, and this annoys historians to no end. If positivist economists have little exposure to the constructivist point of view they have difficulty having a productive conversation with such an historian.
For me, I have no difficulty applying Mundell's trilemma to the past or to the future. But I do realize that the model's deductive exactness depends on a lot of simplifying assumptions. Any person who wants to apply it to situations with pre-modern financial systems had better be aware of those simplifications and have a ready argument (with evidence) that the strictures of the model still make sense.