So putting some more thought into it and looking at the data tables, I think a massive part of the problem here is the entirely inconsistent way of measuring by people.
The first issue is "what exactly is the benchmark for navy balance" because Cybran has a much stronger showcase on traditional navy 1v1 maps due to their frigate. However, their showing on maps like sentons is significantly worse. Likewise, UEF has a much harder time in 1v1 navy but it really plays to its strength in late game navy.
I personally therefore used Seraphim as my benchmark navy and gave them a 3 since they tend to be a comfortable pick for any sort of navy engagement and then gave UEF and Cybran higher ratings because they dominate in their niche. I gave Aeon a lower rating because while they are strong on very specific maps ie sentons, they are absolutely throttled in tons of scenarios due to the sheer lack of frigate AA, including other large teamgame 20x20 maps like selkie or beetle dance. If I instead decided to weigh sentons more heavily than I originally did, Aeon would bump into a 4 easily as a faction. However, I did not but maybe other people did.
However, I see data that seems like people only really cared about senton balance or ladder balance or some other balance. I also see people really just operate on some totally random benchmark. Some people only gave a 4 or 5. Some people only gave 1 or 2. I personally have no idea how you can rate things like this and not have an "average rating" benchmark, but I'm sure if I went and talked to some of the dudes that gave nothing but 4s and 5s we would agree on a lot in naval balance, it's just expressed in a different way.
For me, no faction is a 1 because everyone has a strength to play to and no faction is a 5 because there is no faction that is universally best in any (or most) navy circumstances.
What this means is that the scale range of the data is almost 100% useless. What matters is the difference in values subsets have because that at least somewhat controls for the 2nd issue (could still have a person have the exact same opinion as me but give Aeon a 1 and UEF and Cybran a 5). The 1st issue is just impossible to account for here.