I am an animator in the interactive entertainment industry. My forte' is AI design and character development and how it relates to game design and engineering. I started off as a traditional animator in the film industry years ago... but am now fully entrenched in the world of CG and motion capture manipulation.
Some animators are uncomfortable with the idea of animating to motion capture, as opposed to traditional keyframed animation. In highend video game production, however, when working with realism... your characters movements need to feel authentic and have the subtle nuances that occur in real life. In order for hand keyframers to achieve this we resort to video taping ourselves or others going through the intended actions after figuring out the staging. Every serious animator does this. Then we study our movements and project them onto our character, with added penache.
It is the same procedure with motion capture, except you eliminate the projection on to your character. The data does this for you. I stage the action, don the mo-cap suit, perform the intended scene or movement, and magically (almost) it gets applied to your character. From there I manipulate the timing and the posing to push the performance the way I envisioned and I end up with a stylistic, realistic, nuanced performance that is unique to my vision in much less time.
Anyone who is videotaping themselves as reference to animate to is already doing what is sought after by motion capturing. The authenticity of the movement. But they are missing the time critical middle step of applying the physical data to the model afterwards by arduously hand keying it.
Obviously this technique is not recommended for creatures or more cartoony approaches to animation.