> It only satisfies a weaker condition, i.e., using four non-zero parameters
instead of four parameters.
Why would that be a harder problem? In the case that you get a zero parameter, you could inflate it by some epsilon and the solution would basically be the same.
> In the case that you get a zero parameter, you could inflate it by some epsilon and the solution would basically be the same.
Not everything is continuous. Add an epsilon worth of torsion to GR and you don't get almost-GR, you get a qualitatively different theory in which potentially arbitrarily large violations of the equivalence principle are possible.
That's not relevant here though, because their function is continuous and they're fitting to an arbitrary shape. It's not a "perfect science," so there would be wiggle room.
Why would that be a harder problem? In the case that you get a zero parameter, you could inflate it by some epsilon and the solution would basically be the same.