We would like the option to change the relative weights of the potentials in a combined potential. Example applications are free-energy perturbation
$$\lambda V_1(x) + (1-\lambda)V_2$$
or optimal control
$$ V_1(x) + \lambda(t) kx$$
I think, the best way to implement this is to add a scaling factor $\lambda$ to each potential class whose default value is 1.