The last two or three decades of research in macroeconomics are currently under attack, under the assumption that current research has nothing to say. In particular, the agenda of putting microfoundations in macroeconomics is criticized. Sergio Da Silva goes even as far as to claim that microeconomics need further microfoundations from neuroeconomics more than macroeconomics does. The argument is that even with microfoundations, the Lucas critique still applies (which may be true, if one is a purist, it will always apply), that first principles are not policy invariant (purism again), New-Keynesian models are not well-founded (I agree, but they are a subset of macro), first principles are shaky (there are always exceptions, but for aggregate purposes, they are solid), and micro units are irrelevant because they do not aggregate to a representative agent (but there is also a significant literature in macro with heterogeneous agents, as recently surveyed by Jonathan Heathcote, Kjetil Storesletten and Giovanni Violante. Da Silva recommends to turn to biology for microfoundations and physics for aggregation.
This is very misguided. While I can see the point of understanding preference formation, and I have made the point earlier, cutting back on microfoundations in macroeconomics to replace it with statistical fitting is a serious mistake. Abandoning first principles would bring us back to reduced-form macro, where models were fitted to the data but had little to say in terms of impacts of policy (see Lucas Critique). The current crisis, because it is unprecedented, needs to be studied with microfoundations, so that one can understand how policy will change people's behavior and how this aggregates up. One needs strong theory, not strong empirics, to study counterfactuals in an environment where history is of little help. In fact, microfoundations have recently helped us understand the various policy blunders that followed the Crash of 1929 and led to the Great Depression.