See this question about an origami paper for example. It's a very specific question about a particular theorem the poster did not follow (or in this case the answer did not match that of the paper).
Personally, I think such questions are great. They provide a nice break from all these CW questions we've been seeing. These questions actually have a correct answer, and are usually technical, and indeed it would be hard for the poster to find the solution anywhere else (other than emailing the authors of said paper).
I just wanted to know if the community finds such questions acceptable. If so, we can include it in the FAQ (if we're going to have a list of example questions in the FAQ).