Naked Capitalism has the article Why Central Bank Models Failed and How to Repair Them.
Here is what Yves Smith, who posted the article, had to say to try to make the ensuing paper more understandable to a lay audience.
This is a layperson-friendly description of why the most widely used macroeconomic model, the dynamic stochastic general equilibrium model, is lousy. Most macroeconomists will acknowledge privately that these models are reliable only over a six month horizon, when they are used over much longer time frames. Note that while each central bank has it own version, they all have the same structure and underlying assumptions.
My experience with computer simulation of differential equations helped me understand some of the points of the article. Admittedly, my understanding is hardly complete, but each little piece I can add to my understanding eventually builds into a more complete picture. If you pick up anything out of this article, have faith that it adds to what you know.