Let me describe 2 interesting cases :
Solomon Shereshevsky (Luria, 1968 - the Mind of a Mnemonist) had an almost perfect literal memory. He remembers strings of hundreds of digits for years after only having read them once. I would like to explain this by his awesome synesthesia: everything is encoded in so many ways that everything is considered as new, and so a new memory trace is formed to remember it (we know that "first time" experiences are well remembered (Robinson, 1992), as are distinctive features (Hunt & Worthen, 2006)). Shereshevsky's gift was compensated by big difficulties recognizing (visages, which he considered extremely changing) and had basically no understanding of abstraction: metaphor, figurative language.
Kim Peek, the "real Rainman", had a rote knowledge of 6000 books, yet was unable to extract abstract information from this knowledge (his memory was like a computer's, basically). He was also unable to understand metaphors.
So, it seems to me that abstraction is incompatible with literal memory: we have to forget some concrete features to get the abstract essence. Is there any scientific literature about an incompatibility between abstraction and rote memorization?