Jason: We test our packs on the GFX and attach the voltage sensing leads on the 12 guage wire on the pack side connector. This means the connector isn't part of the measurements. The reason why we do this is we found the tension of the male connector would affect the numbers. Since we cycle 100s of packs a week it's hard to always keep chaging the male connectors to make sure they have maximum pressure. This is also a reason why most of our good racers hardwire and why we don't use inboard tubes. When testing a pack though the Deans connector it adds 1 milliohm of IR and reduces the voltage by .04 at 35 amps. This needs to be taken into consideration.
The packs are charged on Hypeion Duo chargers set at 2C. Then they go onto the GFX for discharge down to 6 volts using 35 amps.
As I have been matching cells for more than 20 years and have tested and been using Competition Electronics chargers/matchers ever since they have come on to the scene in the mid 80s. That being said they try to make the best possible charger andn in the case of the GFX they calibrate them and they do the best they can but it's normaly that they aren't all the same. We use 6 GFXs to cycle our packs so numbers can vary.
The goal when cycling our packs is to make sure they all meet what we feel is the minimum specs for that particular pack. Since were cycling them we decided to post the numbers on the box. So this is a refference that helps and as far as I know nobody else does it.
So feel free to post numbers on your packs when you sell them and I will buy some and see how accurate they are on my GFXs. I think the numbers I get could be different than your as well. This is why you can't compare numbers and when we go to the races we see some racers talk about pack numbers and some get all upset because there packs are lower but it turns out there GFX is reading low and the other guys GFX reads high. Than there is the heat of the pack. When one looks at Humptys numbers he will never get these numbers unless he heats the pack.
As you can see there is no real standards of testing and due to all the variables it is to hard to establish one.