But once your eye has dark-adapted, you no longer fix using your fovea. The point of fixation is shifted 2º from the edge of the fovea after ten minutes adaptation, but drifts back to 1º from the edge of the fovea after an hour of adaptation (reference Y. le Grand's classic Light, Colour and Vision). So you do have rods available at the point of fixation when you are star-gazing. Your highest rod density is at about 20º from the fovea, however (as your diagram shows), so consciously averting your gaze by this amount does improve your sensitivity for very dim stars.
Can you clarify, here? Visible stars have an approximately 1000-fold variation in brightness.* Phil Plait writes specifically about dim stars: "So, while a dim star may be bright enough for your rods to detect, allowing you to see the star, it may not be bright enough to trip your cones, and so you see no color."
So is the reference you're consulting talking about the retina's ability to detect a point source of some specified magnitude (with a luminance figure in cd/m²), or the very different matter of the retina's ability to see by starlight (with a figure for illuminance, given in lux)? If it is the luminance figure, can you let us know that it applies to the visual threshold for stars, about (or dimmer than) mag 6?
(Illuminance from a starlit sky is a more commonly available figure that does touch the bottom of the cone threshold, but isn't relevant to seeing individual stars.)

Grant Hutchison

Edit:
*Not counting the sun!