Can't see how there can be any other idea for grades as long as they're being used as a measuring stick for completing a class, and if they aren't something else will just using another name. You need to be able to know enough about a topic to pass a class, and there needs to be a way to measure that, sure the measuring can be often flawed but unless the class itself if a "learning to learn" class you can't have the goal to be maximize how much you learned about a topic, instead of if you know the topic presented.
> People don't value exam results, in their own right, because they're not that valuable in their own right. A course's job irl is to educate, not accredit.
I agree that the commoditization of higher education is horrible and leading to a decline in something that should be about curiosity and not as a -now, often needlessly mandatory- career stepping stone, but the "free" curiosity part of it should be open classes and such, even in this kind of more relaxed regime you still have limited resources from labs and such that need to thin out their option for candidates and accreditation will come into play sooner or later for those that wish to follow into serious research, particularly in areas where you actually need access to tools that realistically only research labs have. This isn't a huge problem in maths or computer ""science"" but most hard science fields you just can't do advanced research without specialized stuff.