Dealing with this can be tricky, because the answer depends on the question. Often I find questions were written to target an unimaginably demanding goal. These questions often take the form of asking the answer to philosophical questions that philosophers have struggled with for thousands of years to no avail. Or perhaps they take the form of seek an answer that ever military leader in history would have drooled over. In these cases, nitpicking must be part of the answering process because we need to find a way out from the impossible question we've been given.
In less stringent cases, the nitpicking really isn't necessary. I agree that a short question that isn't written in legalees is a much better question for WorldBuilding. No point in making them bulletproof if you don't have to.
Somewhere in the middle, we have the XY problems, where the user really has some underlying need that isn't written into the question. The nitpicks, in this case, serve to try to peer into the outer question that the user really is asking.
The hard issue here is that often the person asking the question doesn't know enough about the topic to know where their question lies. Most people asking about how conscious AI's might operate have no idea that that topic has been studied to death in philosophy, and the answer is "We don't agree on what consciousness is, much less what a conscious AI behaves like." This should be no surprise: they're asking the question because they don't know the answer!
I do think the use of community wikis is the key to solving this. In the case of many topics, rather than nitpicking, we could simply link to a community-wiki answer which is used to provide the same feedback to everyone who asks a similar class of answers. Anyone who is asking for the perfect war strategy in their world can be pointed to a wiki describing just how hard it is to have a "perfect" strategy, and why no general in the history of mankind has ever had one.