NavList:

 I haven't been through the literature on this topic, but I wonder if airglow needs to be part of the visibility equation near the horizon.  We have the "signal" (pointlike star) and the "noise" (airglow, zodiacal light, and other diffuse sources).  It's said that in very dark skies, some eyes can see stars to mag 7.5 or even 8.0.  Without any atmosphere, it might even be a little fainter.

So one might expect a slight penalty in visibility, even after accounting for extinction, because the eye is looking at the (attenuated) star superimposed on a brighter diffuse background than at zenith.  What is the airglow brightness at 2 degrees altitude in magnitudes per square arcsecond, which I think is the usual unit?  One could then pretty easily translate to the visibility limit at zenith in light-polluted skies according to various tables.

Some of the paleoastronomy books have accounts from classical times of barely seeing Canopus from various Greek islands---that might also be useful as upper and lower bounds on near-horizon visibility (with a brighter star).

Cheers,
Peter

ps: do we have any results from ISS astronauts looking through the cupola?  45 minutes is not enough to get fully dark-adapted, but they could wear an eye patch for a couple of hours before trying.  I guess the cupola points mostly in the wrong direction (down), but there still might be some usable sky.