I believe that more than the fact that researchers don't care, is that researchers are evaluated largely by their publications, both on number and on 'impact'.

I believe that more than the fact that researchers don't care, is that researchers are evaluated largely by their publications, both on number and on 'impact'. I just so happens, that the journals with most impact are those that have been long established. These have a track record, a respectable editorial board, and thus researchers try to save up their best work for these types of publications. Disrupting this cycle would require new open journals (or conferences with relaxed copyright policies for their proceedings, especially in the CompSci field) with high quality standards and a respected editorial board. The problem is that there is no real incentive for this to happen, and a strong incentive to publish in the well-respected venues. Hence, the status quo persists...