A recent analysis of plasma stability based on modifications of the extended energy principle for magnetohydrodynamic stability led to conclusions that are too optimistic. The original interpretation of this principle is indeed applicable. The present analysis demonstrates explicitly the fallacy of using the wrong functional for delta W in the extended energy principle. It then shows that the original energy-principle functional delta WB is also obtained for a model in which a surface mass is incorporated to provide pressure balance. This work therefore indicates, but does not prove, that the eigenfunctions that are obtained from a minimization of the extended energy principle with the proper kinetic-energy norm provide a good representation of what would be achieved with an exact treatment.