These estimates retain the circular, edge-on and low-companion-mass assumptions. Adopting amass-luminosityexponent 3.5 instead changes the quoted values by about one percent; luminosity alone does not determine a unique stellar mass without such a calibration.