Why, it's so simple...just a tree-traversal algorithm that treats the TOC as a hierarchical datastructure analogous to XML
See?
Seriously, that's the core of it, a little script that traverses each node of the TOC and can pull out a particular field from each node. You can see how each line has a fixed layout, three fields containing subject line, author, and date; and each field is tagged with HTML to identify it. The date highlighter looks for the "TOCTimestamp" field, the author highlighter looks at the "TOCPoster" field.
So, by way of confirming your speculation, that framework could support script that counts posts by an author in a particular time frame, or pretty much anything that requires taking a look at the contents of each line of the TOC.
Can you think of other kinds of information you'd like to get from the TOC?