After reading every post to this point, I think the root of the anguish and frustration might be related to the mechanism used to shove a listing into one of the three maturity categories. This first pass used a "Pass/Fail" rating system that basically states "any listing containing just one of the many words on the naughty list is automatically considered a naughty listing." Unfortunately, as we have seen by way of numerous examples, this black or white grading system is too restrictive and too inflexible; real life is seldom if ever that clear cut.
I propose that the rating enforcement algorithm be modified to use a "grading algorithm" similar to that used by various Email Spam Detectors. Words on "The List" would not just be used to indicate a naughty listing, but would instead be used to grade a listing. When the program scans a listing, it will accumulate a total grade for each of the two non-general ratings (Moderate and Adult). If a listing receives a total grade that is above the threshold for that rating then it is forced to that rating. The Adult rating would take precedence over the Moderate rating.
This method would not only allow listings with such relatively innocent terms as Gor or Gorean to exist without shoving them into the Moderate or Adult categories, but it would lend itself to smoother, more accurate adjustment as we gain more understanding of "current word usage and meaning". (For example, it used to be totally unacceptable to utter the word "bitch" in public, yet today you hear it quite often and not a soul even hesitates or pays a bit of notice.)
There WILL be exceptions of course, Pamela's birdbath is an absolutely delightful example of this. But I believe a more flexible, less binary grading system will go a long way toward making the system more acceptable to all.