NavList:

Good morning, Modris, 

In answer to your questions,  for determining IE in each of the 6 cases I took 5 readings on, and 5 off, and calculated the average for each, and then calculated as you mentioned.  I have been finalizing my observations with a consistent clockwise turn of the micrometer specifically to reduce the effects of any backlash.  The sextant is a 50 year old Tamaya which pretty much hasn't been used since the '70's, until I obtained it.  So, I think the data is pretty solid.  

Lars suggested that the difference could be due to side error, so when I took a lunar using Jupiter two nights ago I tested for side error using Jupiter and then Saturn.  The apparent disc for Jupiter was big enough that the true and reflected images merged when brought together.... and was hard to actually distinguish due to what I assume is chromatic abberation.  Saturn was better, nearly a point.  Image nearly coincided with the true image.  Though hard to discern, it looked like they were just a hair off of being concentric.  So I would conclude that there is little side error.  But, I don't really know if that difference is significant.  Illustrations of side-error views usually show a wide gap between the two images.