Steve,
I've been following this thread with some enjoyment. The notion of making the observers "blind" to the specimens, unfortunately, won't work. As you pointed out, there are too many identity cues. It would also be arduous, to say the least. However, it is possible to do an experiment, similar to the surveys done by Cornell a few years ago, but with proper attention paid to experiment design and subsequent analysis. For example, if all of the observers were given a thorough questionnaire revealing, age, familiarity with binoculars, ... current equipment, and so forth, it would be analytically possible to tease out some of the bias effects you're talking about. In the end, this would be expressed as a percentage of response variance that is accounted for by demographic factors of interest. In that way, all members of the community could participate, since the analysis would associate response patterns with the characteristics of the respondents.
My guess is that casual birders or neophytes to optics would show quite different survey response patterns than serious birders or optogeeks. Just some thoughts, but I agree with your basic notion that the badge on the instrument would also have an effect that differs depending on the demographics of the observer. It could be very interesting. Of course, it could also be very expensive.
Ed