Does the durability of scientific literature affect research performance assessment at the group level ?