I think you make a great point about how it is not necessary to learn everything, but in a sense it is necessary to know about everything:

Through this time probably 95% of what I've learned will never be used if I become a doctor... but I needed to know that 95% in order to build on it and understand what I will use.

While doing my Master's, there was this professor that was infamous for expecting that his students memorized such things as the exact chemical structure for all amino acids. Luckily for me, he retired and I got the more reasonable professor who cared more about us learning about the properties of the amino acids and their functions within the body, since he reasoned that we could always check the structure in a book whenever needed. In a similar fashion, another professor who taught us cell cycle regulation told us that he didn't care that we knew by heart the names of each molecule involved in every step, but that it was important that we knew how the regulation occurred and which events triggered what action (he even said, "if you can't recall the protein's name, just write something like protein x in the test").

This strategy works because it takes off the pressure when learning and helps people focus in the important principles rather than overwhelming their brains with details that most likely will start to fade as soon as the test is over. To be a good researcher or doctor, you don't need to know everything by heart, but you certainly need to know what to look for and where to find the information you need.

Cheers, and thanks for this very thoughtful comment, @tfcoates!