This paper deals with theoretical predictions for He burning models in a ra
nge of masses covering the so-called Red Giant Branch phase transition. Tak
ing as a guideline the observational constraints given by Hipparcos paralla
xes to the predicted luminosity of models originated from Red Giant progeni
tors with He core undergoing electron degeneracy, we compare models by vari
ous authors as recently appeared in the literature, disclosing sensitive di
fferences in the predicted luminosity. The "solidity" of these theoretical
predictions is investigated by exploring the effects of varying the assumpt
ions about the efficiency of core overshooting or the amount of mass loss,
giving quantitative estimates of the related uncertainties. However, one fi
nds that theoretical predictions concerning the luminosity of the red giant
clump in the Hipparcos sample is scarcely affected by these mechanisms. A
comparison among theoretical predictions as recently given by different aut
hors convincingly demonstrates that the different luminosity predictions ar
e the natural results of evolutionary codes with different - but all reason
able-input physics. In this context observations suggests that stellar mode
ls based on the "most updated" input physics are possibly overestimating th
e luminosity of these structures, raising doubts on several current predict
ions concerning the luminosity of HE stars in galactic globulars.