Comment by pheggs

I dont buy that, mainly because in blind tests where people dont know which model was used and choose a result, this does not seem to hold true