These results are not very promising. They show that these attributes are not to be useful (at least in isolation).
The failure could be for many reasons, such as the fact that they are normalised to a maximum. However, having it not normalised would be just as bad, since then our resolution would not be in a useful range. The other possible cause is that at a frame-to-frame level there is little difference in magnitude changes; and thus the aliasing can occur. Another probable cause is the random noise the glove is generating. It is possible that the major component of the distance covered and the energy generated by a sign are not from the underlying motion, but from noise generated by the glove.