DSGE models used by central banks for forecasting don't do a good job predicting inflation and GDP growth. Even though these models are compared to other forecasts, they still perform poorly. This is similar to how all macroeconomic forecasts have been struggling since the Great Moderation. The DSGE model suggests that inflation deviations should be unpredictable, but this might just mean the model is wrong. So, judging models based on their forecasting ability during the Great Moderation isn't a good idea.