Stack Exchange sites used to have a cookie cutter close reason for "Too Localized". Most of these questions fit the definition to a T. Although the close reason was bumped in favor of more common close reasons, I think the logic still applies.
The idea behind SE sites isn't just to help one person once. The idea is to build a repository of question and answer pairs so that the next guys that come along don't have to go through the hastle and waiting to get answers. The voting system is supposed to vet the existing responses and provide some useful answer to the next person with the same problem. These question are very focused on one person's needs and a thousand factors make in unlikely that the best ranked answers to a previous iteration will serve the needs of the next guy along. Either the information will be out of date or some other factor will influence the details.
I would suggest this metric:
If the questions are about specific details of a specific case, then they are off topic.
If the questions are about the general methodology involved, then they are on topic.
For example, this would be off topic:
What is the closest Kingdom Hall to X?
While this would be on topic:
How do I find the nearest Kingdom Hall?
Another off topic example might be:
I live in X and have a complaint about how the elders in my presbytery handled a situation. Who do I contact to escalate my complaint?
While this might be on topic:
How does the OPC handle escalation of complaints about church leadership above the local presbytery level? How does one determine where to go next?
Note the theme of methodology vs. specific details for a specific case. This avoids both ways these sorts of questions often end up not being useful to the next guy.
- The information gained from answers can be applied to a range of similar circumstances.
- The information isn't easily made obsolete over time.