The way neuroscientists currently think about storage in the brain, it doesn't make any sense to think about "duplicate" data but rather about the "robustness" of a given memory to interference or confusion, which increases with consolidation and reconsolidation of memories or can degrade over time.
Repeated exposure can contribute to the robustness of a memory, but so can other factors such as the emotional content of a memory (which, depending on the emotional state, can increase or decrease memory robustness, and this can also vary with the "jist" versus "detail" of a memory).
Note that there are also very different types of memory, so that repetition is very important for motor plans and "muscle memory" whereas episodic memory can get lost in the repetition of daily recurrent events, such as parking your car or eating lunch: you might remember where you parked or ate today, but if you have a routine this can blur into all one memory, where you have a notion of eating at a particular restaurant many times, but you would struggle to remember specifically eating there on December 14th, 2011, unless there was some other special event on that day.
In your example of reading a book, there would likely be two types of memory: 1) memory for the episodic event of reading the book, which would be a separate memory because it is a separate event, though it could be confused with other similar episodic events. 2) Semantic memory for the content of the book: this would be a single memory, reinforced with subsequent readings of the book, and supplemented with all your other semantic experience with the content of a book: discussing the characters of a fiction book with your friends, learning about molecular genetics from a textbook and from a lecture course, etc.
Note also that talking about the "data capacity" of a human brain compared to a computer is a bit silly, because data isn't stored in the nervous system in bits, and a "fair" accounting of human memory needs to consider the vastly multidimensional interactions that are easily coded in a nervous system but not at all easily in a binary system. Note that other papers have come up with estimates that diverge greatly from the numbers you cite, for example Wang et al 2003 which gives a value of 108432 bits based on the number of synaptic connections. That number is truly inconceivable to our conscious experience even if it is within the capability of our brain. Clearly there are other limiting factors not involved in a simple binary system, for example confusion with similar memories, and there are types of memory that computers excel at (such as the pixel values of an image) whereas human memory deals in abstractions and hierarchical representations.
LaBar, K. S., & Cabeza, R. (2006). Cognitive neuroscience of emotional memory. Nature Reviews Neuroscience, 7(1), 54-64.
Walker, M. P., Brakefield, T., Hobson, J. A., & Stickgold, R. (2003). Dissociable stages of human memory consolidation and reconsolidation. Nature, 425(6958), 616-620.
Wang, Y., Liu, D., & Wang, Y. (2003). Discovering the capacity of human memory. Brain and Mind, 4(2), 189-198.