The exit from a metastable state: concentration of the exit point distribution on the low energy saddle points, part 2