NavList:

>>unless the student is precisely at the center of the dome (sphere) of the planetarium screen, severe angular errors are introduced.  The angle measured is directly affected by the baseline or distance.<<

I observed this problem when I assessed my local planetarium as a teaching site.  I could take an altitude on a particular star, then move to the seat immediately next to me, and get an altitude that was 2° different.

Other issues that I observed:

• There is a certain amount of backlash in the gears/motors that control the planetarium display.  If your curriculum involves using the sky as it would appear on such-and-such day/hour/minute/second, the altitude of stars could be different from one time to the next, depending on whether they were advancing the projector or dialing it backwards.
  
• There was no unambiguous horizon to deal with.  In the dark, it was hard to see the edge of the dome.  I asked management if they could project a laser around the base of the dome to give a clear, unambiguous horizon to work with, and they didn't have the capability to do that.
  
• Our planetarium is 40 years old.  While the dome may have been a perfect hemisphere when it was built, the apex of the dome has sagged by several centimeters over the years.  So even if you could sit in the precise center of the room (which would be awkward, as that happens to be the same spot where the projector is), your results would be inconsistent because of the dome being shaped inconsistently.

I could imagine that - maybe - if you had a brand new planetarium purpose-built for instructing CNav, you might be able to make it work.  But a regular planetarium found in a civic science center is not up to the challenge.  

If problems inherent in the planetarium itself could generate results that could be off by several degrees, then as Brad said, sextant results can be so wildly inconsistent that a student would have no idea whether he was using proper technique or not.

In teaching individuals to use an artificial horizon, I have a couple of times had somebody get a altitude that was 32' from what it should have been.  I could look at that result, and infer that their observation was good, but they were using the upper limb of the sun rather than the lower limb...and were getting results that were one solar diameter wrong.  

Identifying an error in observational technique would be all but impossible in a civic planetarium environment.

If Jon Sutcliff has solved these problems, I would love to dialog with him about just how he accomplished what he did.  I have searched Google, but can find no indication that he published a paper on his techniques.

Bob