Yes, but you have to be EXTREMELY careful and constantly reassess if you are actually asking within the standards. If they start to get flustered because you are asking questions outside what is required, that causes stress and ends up not being fair. It's often hard to put yourself in the shoes of an applicant that studied, was told one thing was most important, looked at the standards, thought he met them all, and then when asked an off-the-wall question, not only is he trying to answer that question, he's trying to understand why it's being asked, if what he knows is wrong, if he was taught wrong, if he should guess, if he should stand up to the examiner (rare), and so on. It creates a snowball effect often that is hard to understand from just the examiner's point of view. I totally agree with questions that actually probe knowledge, understanding, and correlating various systems and scenarios, as long as they make sense. It reminds me of one instructor that attempted to ask scenario questions for EVERY single subject area of any test. He had some highly evolved scenario for each area that had a specific goal and conclusion at the end, but the problem is there are too many opportunities to "branch off", and then you feel like the student didn't meet the standard because he didn't see your scenario the same way you did, or it's so obvious to you because you thought it up, but to the applicant it's not.

The other big problem is that you may think you know a system, but there's always a principle of it that an applicant may know better, so with your highly thought-out scenario, it's possible to trip yourself up. I can't remember how many times an instructor didn't really understand electron flow and grounding, so they really look like an idiot when they try to have some stupid "trace out the electron flow" exercise as part of an oral or check.

The "off the wall" questions must be handled very carefully, and it's likely you can be doing one of the above things without even realizing it.