Good stuff.
When I think in terms of mechanical principles only, and I know this can be a slippery slope, if a humanoid/robot were to be constructed for the sole purpose of linear sprinting from a block start, it only seems logical that low (distal limb) recovery would be programmed subsequent to the start and gradual achieve greater clearance coincidentally as the machine assumes the upright sprint position.
Of course the counter argument to this is that a shorter lever requires less force to mobilize; thus greater heel recovery may facilitate the faster forward travel of the swing leg. I, however, feel that during early acceleration, due to the steepened angle, the low recovery is more favorable.