I'm not a huge fan of scenario based questions, or at the least, I take them with a grain of salt sometimes, because often the "conclusion" or "outcome" that is hoped for by the person asking the question fails to take in all the possible iterations, combinations, external factors and decision making processes that could lead to alternative outcomes. Alternatives that are just as reasonable as what the person asking the question was hoping for, except the pitfall here is when you are asking the question, it is very difficult to avoid "tunnel vision" in the scenario, hoping that the person answering the question is going to get to the pre-determined conclusion that you thought of before-hand. That's the problem with open-ended questions, they may not get to the result you had in mind. Very simple scenarios can therefore be better than more complex ones. IMO, the more complex the scenario, the more of these pitfalls exist and the less objective it becomes. Scenarios have their usefulness, but I've seen them blown way out of proportion as well.

I am a huge fan of actual scenarios, such as when presented in a sim.