How about swapping the highest reading pair in the breaker, and then re-assessing. There must be some sort of R value that is inconsistent, possibly in the new breaker and/or termination ( maybe you could check the heat on all terminations??)