NavList:

 Yes, that's what I get too---azimuth 246.258 degrees, elevation 1.461 degrees.  Or in sexagesimal, azimuth 246d 15.5', elevation 1d 27.6'.  The residuals are around 0.2 arcmin rms.  Maybe an uncertainty of 0.5 arcmin is justified considering timing error and sextant errors.  A third star would have inspired more confidence.

This is the apparent position of the streetlamp under those weather conditions, so using it at other temperatures and pressures in the air path would need a small correction.

Come to think of it, they say atmospheric refraction is extremely small in azimuth, maybe on the order of 10 mas, if I remember right, for non-pathological situations on reasonably flat land.  So some sort of aziumth-only landmark navigation ought to be possible.  More landmarks would be needed, since each one is good for only one degree of freedom.  The elevations of the landmarks would still need to be known to moderate accuracy, but with the right geometry, the high precision would come mostly from the azimuths.  From a terrestrial point of view, this is ordinary horizontal surveying, but then adding some near-horizontal ties to stars.

Cheers,
Peter