Ogy: Where did you get this number? You use the G1 - G2, that is Alnitak - Alnilam positions to calculate it. But at 2550 or so BC, the azimuth at transit of these stars did not go close to 42 degrees or the complementary 48 degrees. The same thing holds for sunrise and sunset. So Don correctly considered this totally arbitrary.
SC: Don did no such thing and neither do you. Why are you concerned with the azimuth at transit of these stars? I am saying no such thing - it's completely irrelevant. As long as you make a good representation of the Belt stars and place these within a square/rectangular frame that results in a line through Al Nitak-Al Nilam (L1) running at 48* it works. You could 'copy' the Belt stars at any point in the night sky onto your plan, rotate your frame until the L1 line is at 48 degrees. It has nothing to do with the angle of the Belt stars at transit. You may perhaps be confusing what I am saying with Bauval's OCT.