Implement local/CI stable test of iseqnormpool picklist contents for metaT notebook

[AmandaBirmingham]

Currently the metaT main path unit test is NOT testing the contents of the iseqnormpool picklist file. This is because the ordering of the records in file is unstable between local and CI environments, which appears to be due to a dataframe merge (a process that does NOT guarantee stable output ordering)--see https://github.com/biocore/kl-metapool/blob/2b2eaece8e5bea2a7588b80d1f438ea19a024e6f/metapool/plate.py#L443C5-L443C20

Unfortunately, changing the order of the records in the dataframe sometimes also leads to changes in which destination well a given input well is transferred to (because there is a running tally of the transfer volumes going into a destination well, and once that well is full, subsequent transfers go into a new destination well. Given this, if you change the order of the source wells, you potentially affect the running total volume for the destination well and thus the point at which it switches over to a new one.) Therefore, fixing this is not a gimme because it requires carefully vetting that the non-trivial change in the picklist created by changing the order are in fact correct.

Blocked by #340