Posting in a seperate comment here as I was mainly aiming to provoke debate with the original one.

This generally chimes with my feeling. It seems to me that the users of research can derive vastly different value from it and so the flat price model doesn't really work.

Consider some new research on light emitting diodes and the value Samsung might get from that vs. me reading it out of curiosity.

For that reason, to me it makes sense to treat academic research as infrastructure and have free access to all funded via taxation.