Gravar-mail: A Highly Efficient Design Strategy for Regression with Outcome Pooling