There are other effects as well, status of the commenter included. People are more likely to read longer comments if the author is known to make good comments. I guess that this effect is weaker for shorter comments where the evaluation of actual content is easier. So, low status karma maximisers should write a lot of brief comments, and as they reach high status, they should swap to longer "walls of text".
Currently, the comment for which I've received the most positive karma by a factor of four is a joke about institutionalized ass-rape. A secondhand joke, effectively a quote with no source cited. Furthermore, the comment had, at best, tangential relevance to the subject of discussion. If anyone were to provide a detailed explanation of why they voted as they did, I predict that I would be appreciative.
Based on this evidence, which priors need to be adjusted? Discuss.