This is a really important study that should be conducted by an organization with a strong engineering culture and as little bias as possible. Who would be better then Mozilla to do this? I really don't trust academia to do this with all of its misplaced incentives.