Problems with the ‘measurement’ of national scientific performance