1. An oldie but goodie: let’s not forget Christina Romer’s classic paper on spurious volatility in the historical data (pdf). This is with reference to recent writings from Eichengreen and Roubini. I don’t favor a gold standard but criticisms should start with this paper.
4. Predictions by Pettis, mostly correct I think, in any case worth a read.