In Appendix A we provide heuristic justification for the scaling of the optimal error rate