Join us at this forum every month for a discussion with famous popular authors from around the world.
uhmm, why would they've calculated the separation between the three points based not on the separation between the three points themselves? Referring to where you say "diagonal across the site and divide by two" as referring to measuring the separation as viewed from the axis of Alnitak-Alnilam. In reproducing a stellar asterism on the ground is it not likely that they would have measured (or attempted to measure) the angular separation between the points rather than using the method you propose, mark a distance (as based on the axis of two of three points) and divide by two? especially considering that the actual difference between the two stellar distances is noticeable, (1min48sec). The difference between the precise belt stars and the precise pyramid placements after everything's said and done is only about 4-minutes of arc per pyramid (if we choose to average and compare like this). This is actually quite good all things considered. It means that; if we so choose to look at it in this way, and to as well divide the discrepancy evenly among the factors, it's like they sighted the stars to 1-arc-min, measured them with a 1-arc-min error from that sighting, set each pyramid out based on that with a 1-arc-min error, and constructed them based on that setting with 1-arc-min of error. Personally; I find the level of accuracy quite good given the scale of the project they were working on and the era in which it was undertaken.