Analogies can be powerful tools in both explaining and misrepresenting a complex topic. What methodology works well for finding good analogies and how does one figure out whether an analogy is well suited for a specific audience?
You can eliminate many inappropriate metaphors by asking yourself a question: "Will the people in this audience understand this metphor?" I can't tell you how many American sports metaphors I've wanted to put into my presentations, only to take them out after a moment's pause to consider my international audience.
Ask a few members of the audience for their opinions.
Finally, I've found that metaphors are awful for persuading. They can be useful to explain new ideas to a generally sympathetic audience. But people who are skeptical of your ideas are likely focus immediatetely in the faults in the analogy.
Analogies should be things that your audience will understand, preferably something that they are familiar with and deal with regularly. For example, I tutor talmud, which consists almost entirely of concepts which are explained by analogy. When teaching kids, I use analogies revolving around sports, and XBox. When studying with businessmen, the analogies come from financial instruments and shell corporations. Just last week I was studying with a programmer so we constructed the arguments in terms of a tree.