Binding Energy of a Nucleus from Mass Defect?

In the previous post when comparing binding energies by RealQM and StdQM for small nuclei, we were led to question the binding energies in StdQM computed from estimated mass defects with the masses of protons and neutrons as input. 

We find that the binding energy of nucleus according to StdQM is mainly potential energy created by the strong force, which does not depend on the mass of neutrons and protons. We thus learn that the binding energy of a nucleus does not depend on the mass of neutrons and protons. These masses only enters in a supplementary computation connecting energy to mass defect including $E=mc^2$. 

We can thus speak about two versions of binding energy:

1. Physical Binding Energy PBE determined by the strong force without input of the mass of neutron/proton.
2. Computed Binding Energy CBE from mass defect determined using $E=mc^2$ to make CBE=PBE.

It may then be tempting in a situation when PBE is impossible to assess either theoretically or experimentally, which is usually the case, to simply replace PBE by CBE and declare that CBE is physical binding energy and not just computed. 

Doing so StdQM delivers a very small drop of binding energy when changing one of the two neutrons of 3H into a proton thus forming 3He, which can be seen as a major change. RealQM gives here a large drop which maybe is more line with physics.