From the perspective of a flight control system design engineer without reference to any specific model, light weight / aft CG and heavy weight / fwd CG usually present the two extremes of airplane response characteristics that we must consider during design. Until recently augmented flight control laws for commercial transport airplanes have been designed without knowledge of the current weight and CG thus the closed loop response characteristics become a compromise between what is found at these two extremes of the weight / CG envelope. Heavy/fwd tends to present the most sluggish response and the highest required control forces as the airplane has its greatest inherent stability at this loading thus requiring more control input to maneuver. Light/aft the airplane is more responsive and has the lowest inherent stability thus can tend toward being overly sensitive.
Another aspect associated with being very light is increased response to turbulence. With lower inertia in all axes a given gust will bounce the airplane more. Add to this the higher sensitivity with respect to pilot controller inputs exciting structural modes and you usually have the greatest tendency for "bio-dynamic coupling" (BDC) at the lightest loadings. BDC occurs when the pilots grip on the controller provides feedback of accelerations that the pilot is experiencing on the flight deck into the system in such a manner that the unintended structural mode frequency inputs that the pilot is providing result in even higher flight deck accelerations. BDC can form a feedback loop that has sufficient gain to become unstable causing structural mode accelerations to grow. Some have viewed BDC as part of the greater PIO topic, but I find if helpful to differentiate between BDC as described here that occurs at structural mode frequencies and does not involve any cognitive, intentional pilot feedback and lower, frequency maneuver control PIO that involves the pilot intentionally closing the loop on one airplane response parameter or another thus creating a closed loop instability. Remedies for BDC and PIO are quite different so it is helpful to make this distinction.
Commercial transport control system design is moving more and more in the direction of sensing / estimating weight and CG for use when scheduling control law gains. This design degree of freedom allows the system to compensate for the differences in airplane response across the range of these two configuration parameters. One thing that we need to be careful about, however, is recognizing how augmentation may be masking the underlying open loop airplane characteristics. This becomes an issue when the weight and/or CG estimates that are being used for control system gain scheduling are incorrect leading to the wrong gains for the current flight condition. In addition, the robustness of weight and CG estimates may not be sufficient to preclude losing them due to detected failures thus having to revert to a backup set of gains or reversionary control laws that do not include scheduling based on weight and CG.
I fear I have digressed from the focus of this thread but appreciate the opportunity to share my perspective on these related topics.
FCeng84