I would say that given the pervasiveness of research on the WCST in academic papers and textbooks, we are not in danger of revealing any information that isn't available to an potential examinee at the public library. That said, a question along the lines of "How do I cheat the WCST?" outlining how to get around the test would probably need to be culled. It's really up to those posting questions on the site to police themselves, as I don't think a formal policy would be very effective, and might smother good discussions about the cognitive characteristics that are being tested.
I understand the point that the neuropsychologists are making on the Wikipedia Talk page, which amounts to the idea that people knowing the scoring system can undermine the test. Parenthetically, I think they are overestimating the damage this could cause. In research, where this foreknowledge of the exam would be most damaging, an experienced investigator would be screening subjects anyway, and in the clinic, there are a battery of other tests and signs that are available to confirm the results of the WCST.
So, in summary, being judicious about what you elect to disclose is probably the best approach. Stack Exchange is a medium for the exchange of knowledge and ideas, so completely censoring talk about what is a critical part of neuropsychological research would be counterproductive.