There is another possible justification for giving higher grades. I pre and posttest all of my students using the Force&Motion Evaluation from Tufts U. My gain is excellent, compared to other introductory physics classes. I can then justify the giving of higher grades on the basis that my students are actually learning more than in other physics courses. The fact that physics is an elective, and is taken by the better students is a secondary concern.
The higher grades seem to have contributed to a larger enrollment, and they have also made the administration happier.
For those not familiar with the gain figures, RPI had about 20% normalized gain <g> with their regular physics course. Richard Hake has inferred by indirect evidence that the HS physics courses taken by premed students produced only 10% gain.
I have also at various times suggested that schools should release the overall grade average for a course along with the student's individual grade on the abstract. Unfortunately this does not take into account that some courses may actually teach more. Unless there are some universally agreed upon standards, grades will only be a loosely approximate indicator of merit. At present education at all levels from kindergarten to graduate school is a very unscientific enterprise. If medical doctors treated patients the way we educate students we would still be treated with leeches for most ailments.
The following web site has a good paper about this problem. www.oslc.org/Ecpn/carnine.html While the particular program being promoted may or may not be the best solution to reading problems, the author's comments about education should not be taken lightly.