NavList:

Art,
Thank you for your reply. I recognize your name as a "guest author" in Bill Morris' blog. I found the write ups on the Kollsman bubble sextant and pelicle replacement very informative. Sextantbook, NavList, and other sites are full of great information. That's part of the reason for my long winded description of my mirror unit investigation...to leave some information that might help someone else.
 
Since I wrote the first post, I revisited the optical modeling after realizing that I could estimate the lens surface radii from Figure 3-1 by comparing the hardcopy to a drafting circle template. Using these radii as constants (instead of variables), I could vary the index of refraction of the glass to obtain the required 49mm object and 23mm image distances in the artificial horizon light path. My best estimates for the lens are  S1 radius 31mm, S2 radius -19mm, center thickness 6mm, Index 1.6, focal length 20.56mm. Changing the index of the fluid in the cell from 1.399 to 1.0 (silicone oil to air) still agrees with the (too short at 15mm) focal length measurement of my faulty mirror unit. I also found that using an optic fiber as a small positionable light source, along with the frosted glass slide to locate image focus, was another method to explore the behavior of the mirror unit at magnifications other than 1:1.
 

I have now removed the two fill/vent plugs and indeed the mirror unit was empty. Thanks for identifying Midwest Lubricants as a potential source for replacement fluid. Before I order possibly 5 cSt and 10 cSt viscosities to blend to the specified 7 cSt called for in the Overhaul manual, I'd like to confirm that the cell doesn't have a significant leak. I'm anticipating some sort of pressure leakdown test with an initial pressure on the order of 1 psig. The plugs are threaded #6-40 UNF, so I'll have to fabricate some fittings first. The plugs have a needle-valve-like tip (.035" dia x ~ 1/8 long) that looks like it was coated in some black elastomer material and enters a similarly sized drilled hole in the port, below the #6 screw threads. This thin coating partially peeled from the tip as it was removed. I wonder if the coating was pre-applied to the screws, or was put on "wet" as the plug was installed. The refill procedure calls for installing the plugs while the whole unit is submerged, so its a puzzle how the sealant was applied. There is no other o-ring or gasket. Today I read some literature about a Krytox thread sealer that seemed compatible (if not exotic and expensive).