Suppose that we have now data sets of weights-at-time for three, rather than two (sufficiently different) known food levels. Which parameters are then estimatable?
