Part of the discussion of the Google Book Settlement is its treatment of orphan works—those where the copyright holder or his/her kin cannot be located—which, by and large, has been cast as a heroic effort to return out-of-print books to the mainstream.
I'm wondering how much diligence goes into tracking down rightsholders before a work is declared "orphaned" (I note the onus in the WSJ article on rightsholders to register works, rather than active searching for rightsholders). My skepticism is fueled by correspondence I receive as managing editor of Clues: A Journal of Detection. Because we publish on both classic and contemporary mystery authors—sometimes the journal is one of the few reliable sources on certain writers and we publish some unique pieces, such as one in the debut issue by John D. MacDonald—and because I have posted on this blog about neglected mysteries and authors, I receive inquiries about rightsholders. One example was in response to my post about The Horizontal Man's Helen Eustis; the individual was interested in turning Eustis's The Fool Killer into a graphic novel. Eustis's Contemporary Authors entry lists a son; a Google search revealed his full name; I found the blog for Eustis's son, Adam Genkaku Fisher, here. Total time I spent in obtaining this information? 5 minutes.
In addition, reprint requests for Clues articles have come my way—initially directed to the wrong people. And these requests come from the supposed digital rights experts.
So perhaps my skepticism about adequate research on rightsholders is justified as we grapple with members of a generation who believe if something isn't online, it doesn't exist and individuals who have a too-casual approach to copyright and are all too eager to profit from the hard work of authors without providing proper compensation.
Also note: This interesting article from the Chronicle of Higher Education regarding errors in Google's book search function.