Using the T-rig produced by the Bake Constraints script, we load up Actions from our MoCap library. We use poses from the MoCap Action to make our own hand-posed Actions. Finally, using the NLA we blend all the Actions together into our final animation.

Since the Actions are bone-specific, we can re-target the armature (scale and edit) to fit any skin, and orient it (location, rotation) to fit into any scene. This video goes with the MoCap tutorial at wiki.blender.org/index.php/Tutorials/Motion_Capture