I suppose the thing to do is find an issue that most skeptics don't know much about, but where there is an open issue with evidence to be examined, yet where expert knowledge is reasonably transparent. I mean, you'd hate to be sitting there while a couple of experts played table-tennis with references:
"Shay says that it's the violation of menis which creates battle fatigue, and the restoration of menis in the primary group environment that permits recovery."
"Ah, but Shay was discredited after his evidence was found to be based on too small a sample."
"Oho, but the sample base for this wasn't the same as the sample base for that."
I mean, that's real sleeper territory.
The chocolate one sounds good to me ... and you can rent those chocolate fondue fountains for empirical testing ...