I believe that is James point, however that conclusion is not taking into account the many different ways the current voters can have voted.
Saying 5 of N is automatically “better” than n of N has a couple caveats: 1. It assumes everyone who voted didn’t do thier research and/or voted for candidates that they didn’t think would fill the position well because they felt thier “extra” votes were not valuable (I personally did not do that) and 2. It doesn’t take into account that voters may have already only voted for the 5 they liked and skipped the rest. My guess is with 91 voters there were a range of boxes checked by voters, some who voted for almost every candidate to some who only voted for one candidate. Some could have voted 1 of N, 5 of N or 15 of N.
My point is it’s extremely difficult (if not impossible) to calculate all the various ways voters did choose to vote this time (because we don’t know) and conclusively say that one method would have produced “better” results than the other.
Also, for the record I don’t have a strict opinion on if there should or shouldn’t be some sort of re-do (I would need to see actual hard calculations to draw a opinion) But I do strongly believe that: