As consistent with my personality, I always attempt to find patterns to simplify things -- and I noticed the same pattern that you have.

I'm of the personal opinion that many people don't read the nodes about questions that they don't know about, and/or don't vote in that section as they're not qualified to determine if it's a good response or not. Philosophical questions, on the other hand, is more a way of stating your opinions, and so you can vote someone up on the appearance of being a good idea, without knowing if it's technically accurate.

I'm basing this mostly on my postings in various SOAP nodes on Perl Monks -- they never get significantly upvoted. If I answer something which I find to be quote minor in some of the more-read topics (mySQL, CGI, OO, etc.), it's likely to get as many upvotes as in the less-read categories that take much more thought and effort.

It's also possible that the technical threads don't tend to get re-read as often. I know that if I read something, and saw that someone had already given a good answer, and I thought it was resolved, I don't tend to go back and re-read it. For the philisophical threads, there is no true resolution, so I will go and check to see what new comments have been added (and thus, I'm more likely to upvote posts that were significantly after the node's inception)

I think the only thing that we can really show is that the number of votes for a node have too many variables involved to be just a measure of the overall quality. Likewise, XP is only an indicator of how much XP you have.