David Brooks produces one of my favorite op-ed columns of the year. Um… what??? Seriously though, he nails it.
The assumptions of classical economics are wrong. You can model a society with incorrect assumptions, but it limits the applications of those models. There is economic research being done using empirically preferable sets of assumptions. Most of it is microeconomic; it is research on individual decision making. We need to further develop aggregated models based on experimental research to further inform our policy decisions.
The biggest lesson: humans are too irrational for markets to regulate themselves. The smaller lesson: classical models on the impact of macroeconomic policy may tend to be right, but should not be taken as dogmatic truth so long as they lean on incorrect assumptions.