There are many challenging aspects of the design of multiple flyby, low-thrust trajectories. One of the most significant, from the point of view of a numerical optimizer, can be the characteristic time scale of the dynamical system. Trajectories in a setting with a short characteristic time scale (i.e. those occurring in the vicinity of Mercury or the Jovian moons) are more challenging to optimize than those with a longer time scale (i.e. trajectories to the outer solar system) because the spacecraft must often perform many revolutions about the central body as well as several flyby maneuvers. This paper introduces modifications that can be made to a multiple flyby trajectory optimizer employing the Sims-Flanagan transcription to improve its performance on challenging problems. These improvements include full specification of the problem Jacobian sparsity pattern and analytical expressions for many of its entries as well as refinements to how the problem constraints are scaled. The improvements are then quantified by solving two challenging problems: a Jovian moon rendezvous and a notional solar electric mission to Mercury.