Thinking that a single system can unify all delicate manipulation capabilities seems to have a low probability of being accurate, say less than 20%, without considering specific application contexts. The baseline rate of manipulation needs varies greatly; grasping an egg is not the same as placing components in a Swiss watch. We should update our understanding that each application has its own critical parameters, like the force required to avoid damaging an electronic chip versus handling a ripe fruit. An optimized system for light loads and irregular surfaces will probably not be the most effective for precision assembly of small metal parts under a microscope.