First of all, as @soandos mentions, there's this from the FAQ to consider:

> However, if your motivation is “I would like others to explain __ to me”, then you are probably OK.

I think this is only part of the problem. The reason why questions not based on actual problems are considered off-topic is to get rid of *hypothetical* questions, or speculative ones. For these, we already have *not constructive* as a reason for closing.

There's a small border between those questions you linked to being off-topic or not. Asking "Why is dial up so slow?" is probably based on an actual problem (having slow internet connection).

I think the way to go about these is to have the community decide. Questions that could be asked:

- **Is the question even relevant for Super User?**  
For example, do we care about how [data travels through the telephone line][1]? Most certainly not. We deal with average and not-so-average computer hardware and software issues. Not more. First, determine whether the question is on topic (users tend to forget this, including me).

- **Can the question be reasonably answered?**  
If the user OP a question that requires a deep understanding of, for example, the ISO/OSI stack, electromechanics, or theoretic computer science, then this is probably out of scope anyway. It'd just provoke extended discussion or inaccurate answers.

- **Are possible answers useful for others?**  
Are they just there for entertainment or are they useful? Do they help to solve anyone's problem? Do they draw traffic to our site for the wrong reasons? Are they going to end up needing to be protected or locked?

At least, that's what I'm thinking about when I see these kinds of questions.


  [1]: http://superuser.com/q/377418/48078