Well, as you say, the degree to which academic publishing is open is currently a matter of some debate. There have been moves, perhaps in some disciplines more

Well, as you say, the degree to which academic publishing is open is currently a matter of some debate. There have been moves, perhaps in some disciplines more than others so far, towards a more explicitly open-access publishing culture at the expense of the journal publishers whose contribution is also debatable these days, but there's a long way to go. What I can say is that as someone who sometimes has an element of research as part of his work for industrial clients, a lot of the useful research material I've read recently has come out of industrial research labs, so they certainly do produce some good work, whatever is also coming out of academia. Another relevant data point: the only software patents that have actually stopped me doing something recently (which I'd come up with entirely independently, BTW) were granted to major US universities prior to their researchers publishing papers.