At BookExpo America, Wet Asphalt’s Eric Rosenfield entered into a lengthy conversation with Brian O’Leary of Magellan Media. And it became necessary to capture their quasi-caffeinated colloquy for reasons that will soon become apparent.

I had seen O’Leary earlier in the year at the “Challenging Notions of Free” panel at Tools of Change, along with O’Reilly’s Mac Slocum and Random House’s director of business development Chelsea Vaughan. O’Reilly and Random House had agreed to participate in a study hoping to pinpoint the effects of P2P distribution — namely, the impact of digital books, both in pirated and legitimate form, on print book sales. And they were standing in a conference room in February to present Magellan’s results to the public.

The results were a bit surprising. According to O’Leary’s subsequent report, “Impact of P2P and Free Distribution on Book Sales,” book piracy wasn’t nearly as ubiquitous as some had suggested. While O’Leary’s report had only O’Reilly and Random House as participants, it appeared that some of the publishers’ fears about piracy were unsubstantiated. Only eight frontlist titles published by O’Reilly in 2008 could be located as torrent files. When these books did become available as torrents, the torrents were uploaded to the Internet far later than expected: some 20 weeks after publication date on average. Furthermore, for the titles available as torrents, on average, sales were 6.5% higher for these books during the four weeks after they were uploaded.

Despite the braying of New York Times guest bloggers, book piracy was hardly the Manichean scenario that some of the DRM advocates had implied. And the chances of Stephen King and Toni Morrison riding on motorcycles appeared to be unlikely. In his report, O’Leary suggested “a less binary model to evaluate the use of free” — one doing away with the parallel experiences from music and movies and accounting for tangible interface realities.

But before the “information must be free” acolytes begin offering a Nelson-like “Ha Ha,” it’s important to note that this isn’t a scenario in which a partisan can dance a jig jig one way or another. O’Leary is pointing out quite rightly that both publishers and open source advocates are making statements about piracy without specific correlative data to draw from. O’Leary’s results are a great step forward, but with Amazon offering a new version of the Kindle seemingly every two months and publishers remaining understandably mum about sales data, it isn’t exactly possible to locate the theory of everything.

In the interview, O’Leary pointed out that not only were there differences in book piracy between fiction and O’Reilly books, but even within specific types of fiction. And getting publishers to participate in ongoing efforts to study this unexamined issue might allow reliable correlations to be formed. O’Leary also alluded to additional studies conducted by John Hilton that involved studying the effect of free digital books on print sales. Hilton was surprised to learn that Tor Books gave 24 of its books away, but saw 20 of the titles with decreasing sales. Random House’s ebook experiments, by contrast, had seen increased print sales for all four titles that it had used for the experiment. But was it the type of books? The specific titles? The way the free ebooks were introduced?

“Certainly when you see that big a swing, you want to look at the type of book or the type of genre or the type of test,” said O’Leary. “I mean, keep in mind that not all digital tests are the same. If you’re using digital content on the first book in a science fiction series to promote the tenth book, it’s different from using digital content to promote the current book. So you want to capture all those things and then start to mix and match over time.”

But with only O’Reilly and Random House willing to use the machines in O’Leary’s laundry room, one wonders if anyone can iron out all the wrinkles.