The check tested f(gamma) >= gamma/(gamma+1), but the manuscript claims f(gamma) >= c1*gamma/(gamma+1) with c1 >= 0.97. It now tests the stated constant and records the achieved one, 0.9711.