With the Mesh Deformer project near completion several people are much more into testing, using, and complaining about the Mesh Deformer. Early adopters are definitely jumping in. That doesn’t mean there are enough test clothes submitted. We still need more. But, those of us trying out the Deformer are bringing up our special cases. Avoid being left out, make sure your stuff works, get involved.
I am interested in using the latest Deformer Project viewer (see the left column for a link to the download) to make tight fitting skirts. I and others have run into problems. That has increased chatter in the JIRA STORM-1716, which I covered a couple of days ago in: #SL Mesh Deformer Discussion. Discussion is continuing.
There are some edge cases that seem to require some mutually exclusive fixes. Adding one type of adjustment fixes one set of problems but, creates or worsens another set of problems. Some of these are extensions of or influences from the basic problems with the avatar that the Deformer will never be able to fix. But, some of the problems are partially from the nature of how worlds built from vertices work.
Give enough CPU/GPU time and programming effort any problem can be solved. But, within RL constraints of time, money, real time render, and render power available to users some compromises will have to be made. The decision process on which compromises is in progress now. I would say the Deformer is going to stay pretty much as it is.
What we are starting to see is some Third Party Viewer Developers figuring out ways to improve the base model the Deformer uses.