Thanks for a good review, this looks like a nifty and useful module.

I'm curious if you've discussed some of your dislikes with the module author? Most module authors I've talked to have welcomed patches and updates. It seems like the hyphens-to-hr (or equal sign) conversion would be a simple enough patch, as well as one to handle /italicized/ text. And perhaps using another character sequence to signal "do not escape html entities within"? Although it seems like the more you add to 'plain-text' processing, the more it becomes its own type of markup :)