Inconsistency in SRRecoParticle::E filling

[mt82]

There is an inconsistency in the interpretation of SRRecoParticle::E between MLNDLArRecoBranchFiller and PandoraLArRecoNDBranchFiller. In the former, the field is filled using the particle’s kinetic energy (e.g. see here),

        if(part.is_contained)
        {
          reco_particle.E = part.csda_ke/1000.;
          reco_particle.E_method = caf::PartEMethod::kRange;
        }

while in the latter it is filled with the total energy (e.g. see here)

     if (m_muonPDG == assigned_pdg)
     {
       recoParticle.E = (*m_trkfitKEFromLengthMuon)[i] + m_mMuon; // [GeV]
       p_mod = (*m_trkfitPFromLengthMuon)[i];
       recoParticle.score = (*m_trkfitPID_Mu)[i];
     }

This issue aims to make the two fillings consistent and properly documented.