Sharing data about library usage, not about subject content

I watched a fantastic presentation on the University of Minnesota's Library Data and Student Success Project.  This project analyzed library usage across 5 types of transactions: book loans, use of library e-resources, use of library computer workstations, online reference transactions, and instructional workshops. Although they had to tie the usage to individual users, the data was only reported in the aggregate. I now have an itch to reproduce their methodology at our library. 

In terms of privacy, they have an interesting graphic that indicates that they had to shift their practice a little outside of the customary library paradigm in which no user identified data is shared.  They are sharing data about library usage, but not about content. So don't worry, Gophers, the library is only sharing the fact that you checked out books, not that you were reading Fifty Shades of Grey.

We kept this:	But not this:
Checked out X books	Actual book titles
Attended X workshops	Actual workshops
Reference interaction	Substance of interaction
Logged into library workstation	Date, location, duration
Used an ejournal	Actual ejournal title

                  (from University of Minnesota Libraries, http://blog.lib.umn.edu/ldss/2012/05/a-word-about-privacy.html)

Do librarians obsess about user privacy to an unnecessary extent? Perhaps. I'll share a true story from my workplace: two of my librarian colleagues approached me with concern, saying that a teaching faculty member had come to the library to inquire whether a couple of his students had been there that morning.  It seems that they had ducked out of a required lecture stating that they had to go to the library to take care of an IT-related task, and the instructor was doublechecking their story.  The request made my colleagues very uncomfortable. In discussing it with them, it became clear that part of their reticence in sharing this information was because they felt that sharing whether a person was in the library would be infringe on users' privacy.

Would most librarians, with our deeply held values regarding user privacy, confidentiality, and intellectual freedom, agree that this type of user tracking is in keeping with those values? Or do our values need to shift in order to realize the potential value for students and our own institutions that can be uncovered by this type of analysis?