Comment on a Monte Carlo investigation of scaling as an alternative to regression