Comments on electron relaxation time measurements from transient reflectances of metals