Good discussants make my paper sound much more interesting and relevant than I had realised they actually were.
Fixed that for you.
Another historian here, but to me a good discussant 1) remarks BRIEFLY on some common themes in the papers, 2) directs the questions from the audience, repeating each question to make sure everyone can hear, 3) fills in any gaps by asking her own questions of the panelists, if needed.
In my field (political science), the panel chair pretty much keeps time (sometimes), directs Q&A traffic, etc. The discussant pretty much comments on the paper.
I just got back from the MPSA meetings in Chicago. Every year, the same old cr@p. Three or four papers presented in series (although this year the presenters were very, very good), followed by comments from the discussant, aimed at each paper with no attempt to try to tease out broad themes (which were ripe for the picking), no attempt to make the papers more than the sum of their parts, which was possible. The way we present papers, at least in pol sci, is tired, boring, and explains why people watch premium teevee, shop, or gossip in the book exhibit--anything but actually attend panels.
What a *good* discussant does: (1) reads the papers (2) draws out common themes and questions for consideration. This is hard work, but you don't earn your name on the program for just showing up);(3) refuses to comment on any paper that is submitted the night before, under your hotel room's door.
Bottom line--offering individualized comments to each presenter, one at at a time, is boring, unhelpful to the panel as a whole, and is not the best use of time at a panel. Try to add value to the panel, not just ask whether some variable's significance is based on a one- or two-tailed test. That's the sort of thing that can be emailed to each panelist.
