From experience on Science Fiction & Fantasy: such questions wouldn't work well, because:
- they are highly subjective: if there are no strong clues in the work then the answers are purely a product of the answerer's imagination and not particularly relevant for other people;
- they are too localized, if they become moot after a future episode/issue provides a canonical answer.
There can be good questions on plot analysis, but I think a good rule of thumb is that if you know that there aren't at least hints in the work itself, then it's not a good question. Answers should be required to say not just “I think X will happen”, but “this documented event suggests that X will happen, while this other event seems contrary, …”
Going on a different tangent, it's too early to ask this question. Area 51 generally establishes what's on-topic and what's off-topic to a certain extent, but doesn't do a good job of setting guidelines for question quality. That's best done during the beta, from actual examples.