It seems to me that you have a good process going, encouraging students to participate and give their ideas. Of course, they are students, so their explanations and insights are likely to be incomplete, even flawed. However, encouraging them is still a good thing and their attempts to put two things together will lead to better thinking as long as you monitor it and intervene as needed.
Your desire to not "shut them down" or discourage them is also to be commended. You do need, of course, to put them on the proper path. So such student comments, like most student contributions, should be treated as an opportunity to go deeper. You can acknowledge the validity of the student's analogy (partial validity) and still suggest that the student also describe how and where the analogy fails. In fact, no analogy is perfect: A is "something like" B is a good way to think about an analogy, not A is "just like" B. In fact A is "also unlike" B is nearly always worth exploring, even for strong analogies. For example, in Java, an object is something like a dog.
Hmmmm.
Well, a dog responds to messages as does a Java object. But I never met a java object that liked to be scratched behind the ears. In fact, the difference is also instructive. Dogs are only somewhat "encapsulated" and solely in control of themselves. You can break that encapsulation with physical force, dragging the dog along by its collar (not a nice thing to do, though). But a properly encapsulated Java object is immune from such things, though you can, of course, re-program the system.
I don't claim my simple example above is definitive, but it was only an attempt to quickly come up with a way to move the conversation in a productive direction.
Some analogies are better than others. But every student contribution is an opportunity for exploration.
