If you have a minimal example of what you would change or acknowledge to improve the post I'm happy to hear it. As it stands, explaining the "why" of zk proofs and an approximation of the "how" seems reasonable to me.
Anyone who doesn't know what a zero-knowledge proof is will give you positive feedback, because they had an easy time understanding the example you presented.
People who do know what a zero-knowledge proof is will correctly point out that the thing you described is not a zero-knowledge proof.
I think this explains the gap you're seeing between the feedback here and the feedback in "less technical forums." It's difficult to come up with real-world examples of zero-knowledge proofs, but there are two good examples on Wikipedia and one here (the "Where's Waldo?" example).
I think my loose narrative could be turned into a proper zk proof with few restrictions- care to give it a try?
The minimal example of what would have to change would be something along the lines of this: rather than asking specific questions, you would want the stranger to be able do something that only a family friend could do. But, in this case, it's silly to get them to do that when what you want is that knowledge -- who they know, how they know them, etc. -- so you can verify it against what you already know. The question of whether or not someone is a family friend is just not the type of question that these proofs are used for.
I would throw out that example and use something like this: a lock box that anyone could close -- but only someone with knowledge of a secret code could open it. Someone could prove that they know the code by simply opening the box. It's a contrived example, yes, and could probably use some extra pizazz, but I think that's because you'll find most real-world situations simply don't work well with these sorts of proofs.
Also, is somewhat relevant: https://news.ycombinator.com/item?id=14261067