Charles,
    In effect, what you have described, is to take advantage of the same property that is used by the folded pendulum, which
comprises both a `regular' pendulum and also an 'inverted pendulum.  Separated from each other and connected by a rigid
horizontal boom, their relative influence ('restoring' from the one, and 'destoring' from the other) is determined by how close
the inertial mass is placed to one or the other.
    Because the folded pendulum can be made to have a very long period, upper valuve being limited by mesoanelastic complexity,
it appears clear then, that the feedback drive of the primary pendulum by an inverted secondary one is capable (for ideal
meaterials) of very long period indeed, and therefore very great sensitivity.  Moreover, since the adverse effects of material
problems can be essentially eliminated by means of the feedback, I see this as a really attractive idea to try and demonstrate!
Are there any takers?  (meaning folks like Brett who know how to make control systems work right).
    Randall