In this question a provided answer is very detailed and informative, but due to a misunderstanding of the question potentially irrelevant. It would be a shame to lose that answer and delete it. How can we retain the value of the answer without misleading readers of the question?
What we've done in the past has been to include some disclaimer at the top of the answer saying that the question was misunderstood, and the answer corresponds to some different interpretation of the question. A good example of this would be the question Drawing samples from a finite mixture of normal distributions? where my answer and Barron's answer both corresponded to the same misinterpretation of an earlier revision of the question.
As for the original poster of the question, if people are misinterpreting your question, you may want to think about revising the question to clarify any ambiguities that have caused the misinterpretation (such as terminology, unclear wording, and so on).