Any of these options are things a writer may do and may find beneficial in their particular circumstances, but I find choice D to be most important. Like choice B, it emphasizes that reviewed papers are a useful source of self-examination and noticing what went wrong, but the language of choice D is more neutral and less punishing, encouraging the writer not to beat him- or herself up for past mistakes but rather to merely be heedful of them and learn for them so as to improve. I would be inclined to add that keeping a collection of reviewed papers can also allow you to revisit what went right and learn from successes in order to replicate them, especially if the write is on his or her way to acclimation in a new environment and must learn the tastes of a new professor, boss, institution, etc. Choice C is also quite a valid approach to learning but seemed slightly less relevant to the question at hand - that is, a physical copy doesn't seem totally necessary for remembering a second opinion, while that physical or digital copy may be more helpful for recalling less broadly memorable issues, like technical or stylistic errors - and choices A and E just didn't strike me as as important for success.
