I do not think the question-first strategy is the way to go with any section of the MCAT. I do not want to give too much advice on VR because I am struggling with that section quite a bit, so I will keep this post related to the sciences.
I've done numerous passages for the MCAT science section and one thing I noticed was that when I first started out, I was very uncomfortable with the test layout, and tended to use the Kaplan method of mapping the passage. Now, I must admit, I used this technique incorrectly, because I was focusing on the details quite a bit and writing all of those down, instead of thinking about the topic at large and really trying to make sense of the point of the experiment. I wasted a lot of time and as time went on, and I did more practice problems, I noticed that oftentimes, the passages can be loaded with information, and very little of it will be featured in the questions. However, that is why you need to practice, and get used to this idea, and over time you'll become quite familiar with what information is useless and what is not when you read a passage. I do tend to map a passage in the sciences, but I keep it extremely, extremely brief. I'm really just jotting down some notes more than anything else.
I find it helpful for experiments where two methods are proposed or passages in which two theories are proposed about a certain biological/organic mechanism.
The question-first strategy ironically will actually keep you too focused on details and not looking at the bigger picture, which one needs to really succeed in any section.
Long-story short, I do not advise the question-first strategy at all. The passages tell you a story so you want to try to visualize the information and imagine it actually happening to get a more "global" picture. When you read the questions first, you remove that aspect and make it harder for yourself to understand the actual point of the passage and will find it difficult to structure the information. Also bear in mind, that among other sorts of "trickery", AAMC testmakers plant answers that are simply reiterations or rephrasings of what a passage said, but are not applicable to the actual question asked. This is in a sense meant to bait individuals who simply look for something familiar instead of thinking the question through.