As someone who works on assessment and learning in higher ed, this - grade inflation and snowflakery - is something I deal with all the time. Couple of things that haven't yet been raised.
1. Grades are, at best, an incomplete measure of learning (and sometimes - too often, if you ask me - a complete MISmeasure of learning). Just so I'm clear, here are scenarios. I assisted a colleague in making his course more student-centered, which sounds touchy-feely, but all it means is that we use data and evidence to tell us whether our students are learning, and how much. Student-centered instead of instructor-centered - that's the model a lot of us grew up with - sage on a stage, instructor droning while students do nothing but take notes - still with me? Anyway, what my colleague found out from using a variety of other measures was that the A students in his class didn't really know much more than the C students - they were just better test-takers. In fact A students still left his course with some really major misconceptions about their subject. Over the span of four years, he and others at my uni have reformed their courses so that their assessments - quizzes, tests, projects, papers - are much better at telling both students and instructors whether their students have learned anything, and what they've learned. Ironically, his DFW rate (rate of Ds, Fs, and withdrawals from the course) have dropped from ~45% in his introductory course, to ~17%. (Forty-five percent is educational malpractice if you ask me.) Meanwhile, more than half his students earn - EARN - B-minus or better. Typically 20% earn As or A-minuses. His students are more likely to persist as STEM majors, too. This is in an introductory large-enrollment science course that's a feeder for all STEM majors, and because it's used by ABET in their engineering program accreditation, he must still demonstrate that his students are learning what the institution and ABET expect. He can do this in spades.
2. I've been doing this "merely" ten years - much less time than MAJ Hanak. Still, in those ten years I've seen the snowflakery increase exponentially while skills like staying on task, persevering with difficult problems, focusing on ONE thing at a time, self-motivation, and follow-through have declined precipitously. That's the legacy of NCLB: teaching to the lowest common denominator instead of pulling the entire population along, deadfully boring drill-and-kill test prep for assessments that don't measure what we really want from our students and poorly measure what it is they've learned or not learned. There are studies beginning to emerge about the specific failures of NCLB for almost everyone. Suffice to say that the freshman year of college is fast turning into a crash course on how learning works in the REAL world, kiddos, and it ain't pretty. These insufficient life and learning skills are NOT restricted to the average and below-average students, either. In fact (anecdata now) I see "smarter" kids having the laziest attitudes, which makes sense if you think about it - if you're smart, and it's always been "good enough," why push yourself?
Alas, Harvard is just a symptom of how dis-associated grades are from actual learning. But there are people like me, at many institutions, who are working on how to motivate our students, prepare them soundly for STEM professions and to participate as citizens in a world where we have some very difficult decisions to make as communities and societies.