In order for macroeconomic research to be successful, it must be able to
The latter seems to be the primary goal of most macroeconomists, but I don't think that's particularly helpful unless you can do the former. Macroeconomics has to be helpful to central banks, governments, firms, and consumers in order to have value, and that isn't possible if you are only able to look at the data after several years have passed and tell a story about what might have been going on.
I'm not saying macroeconomists should have predicted the financial crisis. It would have been nice if they had (in which case it might have been prevented). Ultimately, there are unpredictable shocks, some of a very large nature, and there's nothing that can be done about that. Once those shocks occur, you call in the macroeconomists to explain what's going on, and they tell you what outcomes to expect conditional on particular responses of policy and the private sector. You don't blame the plumber if you have a water leak in your house, but you expect to be able to call in the plumber to diagnose the problem quickly, and follow that up with offering some solutions.