Benzylpyridinium salts (BPs) have often been used as thermometer ions to obtain an energy calibration of mass spectrometric experiments (in particular to determine internal energy distributions of ions after the ionization process). Fragmentation of BP + molecular ions is characterized by specific Bond Dissociation Energies (BDE) which depend on the substituent group and its location on the benzyl ring. Although those BDE values are regularly re-evaluated by quantum chemical calculations, their experimental determination is still missing from the literature. In this paper, a modified Quadrupole-hexapole-Quadrupole (QhQ) mass spectrometer is used to obtain such values on 4 BP + molecular ions (characterized by a wide range of C-N bond strengths) using Threshold Collision-Induced Dissociations (TCID) and Rice-Ramsperger-Kassel-Marcus (RRKM) kinetic modeling. It is found that experimental values are systematically 0.5 eV lower than their most recent theoretical evaluations. Despite this shift, the absolute critical energy values are maintained in the same order (pOMe < pMe < pCl < pCN), and the relative energy differences are in very good agreement. We argue that the observed 0.5 eV shift relates to the energy dependence of the Transition State's number of states that is typical for barrier-less fragmentation processes and, relates to a kinetic rather than an energetic bottleneck. Notably, by taking into account the bond elongations characterizing the transition states and their corresponding calculated critical energies (E0), close agreement is found with experimentally obtained E0 values. We thus conclude that much care should be taken when describing the transition state during an internal energy calibration procedure where the involved energy is lower than 3-4 eV.