I feel that when I ask a leading question it generates a deluge of content that is hard for everyone to incorporate or engage with.\nI'm not really sure what kind of leading questions that you ask and how it steers off, but it's OK if this happens. What to avoid is to give this too much space if you can't continue. Let it instead go and find another subject to talk about. It's basic conversation principles. Why, for example, strangers-that-meet-on-social-events ask what they do for a living, it's to find a common topic to talk about. If it doesn't bear fruit, drop it and continue on with something else. Imagine how awkward it would be between two people if someone were persistent about the subject, for example work, if it didn't generate a common ground to talk about.
The garden variety leading question in a story game is something like the following. "Mr. NPC hates you because of an interaction you had 10 years ago -- what was it?" or "The giant purple tentacle knocks over something dangerous -- what is it?" In my experience, if the question doesn't call for a lot of exposition, it usually won't throw the whole game of course.\nLeading questions comes in at least two shapes, and you have to ask yourself why to use it. The examples above can put the player in the spot. I've done it to other players and seen how they froze up for a couple of seconds and I have had done to myself as well. It's not a comfortable feeling to be pressured like that. I do find the fun in playing with descriptive challenges. Heck, I'm currently producing This Is Pulp that's all about that, but you still need something that the players can lean onto when you do that. The questions above doesn't have anything to leave. You just shove it onto the player's hands to handle it. It's like that old joke in improv: "And the next thing is a really great idea. Simon, tell me about it."
Archipelago does, but the other players can say STFU at any time.\nSee, that's why I like the earlier edition better.