The other side
Jaime Ros, with Pedro Hugo Clavijo, wrote a critique of Thirlwall's Law (in Spanish). Replies by Carlos Ibarra here, Esteban Pérez here and myself here. Their rejoinder here. All in Spanish. Haven't read the whole rejoinder yet (just got it), but for some reason they insist that the supermultiplier implies that exports are always the main source of autonomous demand.* Hm, that's weird. Just puzzled by that one. They should read Bortis and Serrano (this one with Fabio Freitas). Oh well.
* In their words: "El problema con esta propuesta [el supermultiplicador], ... implícitamente está suponiendo que las exportaciones son el componente principal, aunque no único, de la demanda autónoma..." Translation: "The problem with this proposal [the supermultiplier],... is that implicitly presumes that exports are the main component, even if not the only one, of autonomous demand..."
Gets better. They agree with me on this. In their words: "estamos de acuerdo con Vernengo cuando dice: '(…) el modelo de Thirlwall no es capaz de captar la complejidad del proceso de desarrollo asiático, el papel del Estado, de la inversión pública, del financiamiento de los bancos públicos, del papel jugado por factores geopolíticos asociados a la amenaza comunista, entre otras cosas'." Translation: "we agree with Vernengo when he says: '... Thirlwall's model is not capable of explaining the complexity of the Asian development process, the role of the State, of public investment, of public banks' financing, of the role played by geopolitical factors associated with the communist threat, among other things'." The rest of the paragraph they forgot to quote: "Several of these factors were operative through autonomous government spending. One cannot think about the Chinese case without public investment." That would be as in government spending is the main, even if not the only, source of autonomous demand.
PS: The whole issue of Investigación Económica is available here.