The biota doesn't seem to end up in the same place as the biome. I think there's a bias moving things in one direction or another. Small sub-biomes sometimes end up entirely devoid of their biota. Then, sometimes, trees can appear far into the "wrong" biome. One valley sub-biome (of Highlands Cliffs) had a strip of tree-free Valley on the south side and strip of treed Cliffs on the north.

Sounds like this would explain the underwater forests and clearly ocean biomes showing a land biome name instead. :)

Underwater forest, could be. Ocean biomes showing a land biome is a different problem - land near ocean can end up submerged in large tracts because a) some is supposed to be "lake" and b) smoothing. I will be able to "repair" some of that in the future using the same techniques in the upcoming snapshot I use for the beach but right now I'm also having trouble with river conversions and I need to see what's going on in the "water zones" to fix that.

Could it have something to do with this:

Always wondered what that line was actually doing.

Yeah, that could explain why the bias goes in different directions (I think). I'm guessing maybe it's to keep us from drowning in decorations? Research project for an upcoming snapshot, I guess

Fixed in 9574ab3