Analogies are sometimes useful for understanding, but they rarely seem useful in marking an argument. It always seems redirect the argument into the ways that the analogy is different. (If the analogy weren't different, then it wouldn't be any clearer.)

If one goes into it with the intention of using the differences in the analogies to illuminate the domain, it can be helpful. But any argument of the form "X is like Y, and therefore since p(Y) is obvious then p(X) must also be true" is pretty much destined to shed little light but much heat.