My understanding is that Eric Flint's free library experiments are focused on the giving away of electronic versions of older back catelog works as a way to find new readership. He's got credible numbers to backup the positive bottomline effect for book authors. But a lot of that positive economic feedback is about deriving value from older works. In Eric Flint's experiments authors are maximizing the value of older works and getting better royalty paychecks as a result.
The conclusions that can be drawn from the Baen Free Library experiment are not directly applicable for the computer software market in general and certainly not the linux operating system market aimed at business end-users like the ones at the LF End User Summit.
The market dynamics for operating system software are intrinsically different than what fiction book authors are presented with. For books and music.. older works still hold a significant amount of value for both the producer and the consumer. When a new reader "discovers" an author there's a reasonable chance that new reader will want to experience some or all of the authors older works...not just the latest book. That doesn't happen for linux distributions or for most categories of software. If your first experience with OpenOffice is version 2.0 there's very little chance you'll choose to go back and use OpenOffice 1.x unless your forced to for some bizarre technical reason.
By and large software releases are about incremental improvements... not completely new works... and thus the back catelog ecnomics just don't exist. The only area of software that has the same sort of back catalog value dynamics which drive the economic feedback that makes the Baen Free Library worthwhile for authors are software games titles.. and that's pretty much out of scope for the discussion at hand. We aren't talking about unpaid deployments of software games. No for operating system deployments we don't have any metric driven picture of how unpaid versus paid deployments really impact the market... and Eric Flint's metrics for book authors participating in free library activities don't map over.