And any reasonably smart person could sit through a bar review course, do the work, and pass the bar without ever having set foot in a law school.
And any reasonably smart person could pass most final exams after attending a well-conducted two week review course based on the actual content of the final (as is a lwa review course) without ever attending the class in question. So what?
The important relationship would be between measures of success in law school (measured by things like membership or even editorship of Law Review, overall grades, grades in courses in one's current speciality, etc.), the quality of the school, and success in the profession. My observation (very limited sample I admit) is the results of this prediction model would be stronger than you think.
I don't know what field you are in, but I seriously doubt most people could pass most final exams with a two week review session without taking the class.
Sample Question: Part 1. Verify that the 2s AO 2s(orthogonal) = (1-S)^1/2•(2s-S•1a) is orthogonal to the 1s AO and is normalized.
Part2. Let an MO have the form a(1s) + b(2s) + ... when expressed using a nonorthogonal 2s Slater Type Orbital and the form c(1s) + d(2s)+... when expressed using an orthogonal 2s Slater Type Orbital(from part I). Show that c=a + Sb and d=b(1-S^2)^1/2 where S is <1s|2s>.
Yes, not too hard if you have had the class, but really difficulty without the class and the prerequisite math classes.