Consider the quantum coin flips as a branching in the wave function between worlds in which Contestant A wins and worlds in which Contestant A loses. Under the previous understanding of the QRR game it didn't matter what the probability of winning was. So long as all the worlds in the branch of worlds in which Contestant A loses are unoccupied by contestant A there was no chance she would not experience winning. But as soon as a single world in the losing branch is occupied the probability of Contestant A waking up having lost is just the probability of her losing.

Given QI, we declared p(wake up) to be 1. That being the case I assert p(wake up having lost) = (15/16 * epsilon)/(1/16 + epsilon*15/16).

It seems to me that you claim that p(wake up having lost) = 15/16. That is not what QI implies.