DSGE and ABM

Solow explains to the Congress the problems of using DSGE models in economic policy decisions. The Economist picks up from here and argues rather to use the well-known alternative approach of bottom-up agent-based models. Is it true that computational tractability is the only thing that's hanging on us? Will we reach a concensus on the behaviour of each economic agent and their interactions? Seems like an interesting field in the future.