cool idea. Let me see if I understand it correctly...
I want an animation from A set in B set, but B set has a cape, or hidden guns or something.
With the program, maybe duplicate an animation from B set, that has the cape nodes in the keyframe data, then copy the A set keyframe data into that duplicate animation from B set?
Or...
A set has extra nodes in the animation that the mesh for B set doesn't have. The program will read B set's keyframe data to make a list of all the nodes that are used in B set, and when you copy over an animation from A set, only nodes present in B set will be copied over.
I don't think the first scenario is needed, as mesh pieces will just stay static if no animation data is in the keyframe file. Although it would be beneficial for hidden weapons to stay hidden.
In the second scenario, the biggest issue I see would be with the extra data. There are a few ways I've dealt with it in the past. The more time-consuming way is bringing all of the relevant information up, and then deleting the extra keyframe data. The easier way is just renaming the extra nodes to something that is already present in the keyframe set and referencing those animations. That's why if you look at some keyframe sets I've done in Nifskope you'll see Head node animation data 3 or 4 times, I used it to cover up nodes that didn't exist on the mesh without changing the file size of the animation.