Minimax Aspects of Bounded-Influence Regression: Rejoinder