Since getting a VR rig 2 years ago, I found a program in beta called Tvori. Easy animation with either real-time hand waving or frame-by-frame. I think they are heading in a Enterprise direction so they really should put out a kid version.
I've also been invited to beta test Animotive.
Sans VR rig, Blender would be the one to try.
I honestly don't know how to convert their XYZ dxdydz into a Eshkol Wachman notation like stickman in hamster cage, but that's what programmers are for, of which I aren't for a long time.
But it also raises the idea if all you want is motion capture data, of which there is a shit-ton and a lot is fight stuff for games, that is an easier route than homespun karaoke.