Dynamic Generalized Linear Models and Bayesian Forecasting: Rejoinder