Bayesian model selection based on proper scoring rules - Rejoinder