" ... how is it that we can remember past incidents, yet don't connect with the similarly discrete version of ourselves having that experience?"
Because there is no discrete version of ourselves.
Like Bell loyalists, you mistake the discrete processing of information in your mind for continuous objective information independent of mind. Remembering is a selection process -- you choose your memories, i.e., what sequence of events makes sense to you, or aids your survival.
The choice function in nature is indifferent to the choice function of what Gell-Mann calls information gathering and utilizing systems (IGUS) which we and all other organisms share in common.
As Rob implies, a wave guided by a landscape characteristic (a slit) is like a simple version of a brain-mind adapted to the configuration, such that the resulting information is a product of the configuration space, rather than new information created by a brain-mind observer.
Rob's model fails at the level of completeness. That is, the configuration space is covariant with the observer; nonlinear random input that smooths the function of mutual interaction of IGUS with configuration space, is continuous. The function is also self-similar at every scale, so that complete information is manifestly local because there is no local-global boundary.
The question, then, of how many "you's" exist on a continuum is the result of naive realism. The configuration space is never nonlocal, such that would allow one to count selves on a continuum of selves -- the point at infinity is also local. Therefore, only the correct topological configuration -- as Joy has demonstrated -- guarantees the global self-similarity of a unique "you," independent of what you think.
Best,
Tom
