Backlash & a lack of stiffness have the same effect, but the cause happens at different scales. Just because the gaps are in between the atoms of the dyneema doesn't mean there are no gaps! There's still play due to a lack of stiffness, just less of it. The positioning errors should be lower than the feedback resolution on the servo encoders, but they won't be zero.
I disagree; backlash is unrelated to force. If you were to graph error over force for backlash it would be a straight line but for something like dyneema it would directly scale with force applied.
In a driven system the dead zone is graphable and that deadzone time or area does not change based on velocity, A lack of rigidity in couplings, gears or belts IS affected by velocity