A quick search by way of Google Scholar reveals that worldwide quite different grading "systems" are being used. Well, the term "system" is too much honor. It's no more than a local (e.G., USA) convention to *report* about the performance of learners on school, college or university. This kind of report is very coarse, especially when using only five (A,B,C,D,E) levels of performance. Appending qualifiers like plus's and minus's doesn't change that essentially. Although whatever you will propose and use instead is likely to be criticized on the same or other grounds, I would rather prefer more symbolic forms of communication about learners' achievement level. If learning or performance assessment is based on a sound scoring system based on psychometrics or other quantitative approaches, why using letters at all? Numerical grades have their problems, too. One problem is that people may think that those grades have other properties than denoting a level on ordinal scale, e.g. falsely believing that you may build averages of such scores. Another potential problem with them is that they suggest an accuracy that you can't ever arrive at. Still, they have a built in order relation, which is absent for letter codes (is A greater than B!?). Multidimensional iconic codes like Chernoff Faces may be the best solution, after all, because they are very intuitive and don't invite you to project too much on them.
The letter grade system is successful in showing what students might lack when in compared to their peers. But, do we really want that? I agree that an accurate assessment of areas the student can improve on instead of showing which areas his/her peers are better in. (Although the letter grade system does try to do that, it doesn't really achieve that) I feel that such an assessment will reduce that bloodthirst in educational institutions to a healthy competition. After all, the only thing that matters is what the individual has gained from education not what his/her peers have done better in that same period of time.
Yes, there are any number of useful alternatives: standards-based report cards, for example maintain a high standard and give a more accurate assessment of what students know about specific aspects of a content area. This is a significant improvement over a simple letter grade that doesn't tell me about both the strengths and weaknesses of a learner within the course.
When parents get their child report card, they do not know what it means when they get "proficient" in math, they understand percents, and letters. It is the system we all know and can understand. By changing it parents will not be properly informed about how their child it doing academicly.
The system (A, B, C, D, F) is not a bad system, and even if we replace the letters with numbers, the basic structure will remain. The problem lies not in the reporting system, but in the way the grades are generated.
For example, tests, homework, quizzes, and projects earn you a specific letter grade. If you are a great test taker, you generally do well in the class. The problem is, for those not-so-great test takers, your grade reflects that. It isn't your knowledge or even lack of knowledge. It's the test.
Perhaps if we were able to structure the system to grade based on curriculum standards (For example, in a math class, Being able to properly apply math operations for various equations). You have to learn the skill and apply it regularly. There are several ways to assess that skill that are not test-specific. Rote memorization is almost necessary in school today, but rote memorization is not learning.