[Optimal Designs for Comparing Test Treatments with Controls]: Rejoinder