Figured out what happened, one of the alligator clips, the wire strands inside broke off, didn't realize it and only 1-2 tiny strands were holding it. So the charger didn't really charge module 23 that well because a lot of current would of been lost as resistance through the poor condition.
Welp, that didn't fix it. In fact capacity fell further from 3194 to 3069, so I think this module is toast.
The rep for hitec usa has yet to reply to my interrogatory on this charger. The manual states the charger can determine internal resistance, but it does not state the method used, which affects the interpretation of the results.. Is it injecting an AC voltage of some frequency eg 1,000 cycles and using that or Is it using the DC load method with the charger placing one or more loads on the cell for a second or so then executing math on the resulting voltage readings to determine the resistance both methods are valid, but yield dramatically different results, one must know the method to interpret the results Thanks