Offhand, if you're comparing ONLY the RG, I'd use the average. Since the RG differential (diff of max and min) is usually different, the max or the min doesn't give you a true picture. However, ..., ball park estimates are usually more than enough. Example: the difference between 2.45 and 2.55 is significant, while the difference between 2.52 and 2.54 is not.
The idea is:
Yes, in case #1, a 2.45 RG ball is going to start revving up long before a 2.55 Rg bal, given the same cover (gotta keep some variables fixed).
In case #2, No, I doubt if we'll see any difference between a 2.52 RG ball and a 2.54 RG ball, given all other variables remain fixed.
Besides, once you drill the ball, all that will change; the starting (undrilled) RG is good enough in the 95% case.
