I'm sure this is well known to most of you but I'm going back to basics on how to choose cell sizing to get fastest speed. Following Harris/Weste 4.3.3: choosing best number of stages