An Observation

Interesting stuff. Not sure why it's on the wiki and not the board, but this discussion is something we absolutely must have.

Although I'm not a biologist by discipline and have no formal training in the area, I have managed to pick up some knowledge through trying to design aliens, dragons, etc.

The biggest problem with a Suvian classification tree is that in World One biology, organisms descend in a treelike manner from common anscestors. The products of bad writing don't. However, biology can handle convergent and parallel evolution with little trouble, so the resulting duplication of traits across artificially-imposed tree boundaries isn't a big problem. Despite the enherent problems with forcing a tree-like classification onto something that's not tree-like, I think we should still make an effort to clump Sues by shared or similar traits. The more broad the similarity and the more variation is observed within the group, the higher up the tree it should go.

I would start by seperating Sues into two kingdoms (or perhaps a new, higher designation): wraiths and physical Sues. I'll look at the Wraiths at some other point (focusing on Sues now), but it seems obvious that given the massive physical differences between the two, they should be very far apart on the tree.

Sues I'm thinking of seperating at the next-highest level into "active" and "passive": roughly, Stus and Sues. However, I'd like for this not to be based on biological sex, as that is in and of itself relatively superficial. Amazon Sues will fall into the "active" camp (assuming they are actually competant, otheriwse they are an offshoot of passives), while wimpy Stus (such as Archir the Ring Child) will fall into the "passive".

I feel like this needs a flowchart.
I'll write more once I figure out how to make & post one.

Ok. I'm proceeding in a roughly depth-first method through this tree, ignoring those sections that are more difficult to classify with the idea that working on the easier ones will give me more insight that can be later reapplied.
Tree can be found here:

Primary reason for considering a seperate category for character replacmenets is to make the tree look a bit less binary, and that they are often physically different from Wraiths or corporial Sues. However, since character replacements can do MANY different things, it might be better treated as something like camouflage, evolved independantly so many times that it is useless as a classifier.

Decided to group the passive Sues by thing they are intended to elicit. These roughly correspond to the suetypes from TVTropes:
purity sues (including literal goddesses, druids, etc.),
relationship sues (a subset will be porno-sues whose only involvement is to give or recieve sexual gratification),
popular-girl sues,
villain sues (those not actually badass enough to fit into the active category, which seems to be all of them?),
sues that primarily elicit pity or receive hurt/comfort (providers of comfort are a purity subset),
and sues whose primary purpose just seems to be showy and pretty (maybe merge with "decorative"?);

I've also branched off at this point:
sues that exist to convey an ideological message,
"false badass" warrior sues,
anti-/suetiful-all-along sues,
and probably a catch-all "other" category.

Not sure if trans-dimensional hoppers and snatchers are broad enough in function to merit their own categories, of if enough of them fit into the relationship category to justify making them a subset.

Considering adding a third "peripheral" category in addition to active and passive, for Sues that don't directly exist to represent or act as a vehicle for the author: love interests, villain sues where another sue is clearly the authorial focus, deus-ex-machina "cavalry", etc. Could make these another type, but they can be either active or passive in roughly equal number.