The idea that a simple capability demonstration with a red ball can directly translate into optimal grasping of an apple underestimates the complexity.
To achieve state-of-the-art performance in fruit manipulation, sensors that detect asymmetric pressure and surface are needed.
Handling an apple without damaging it requires a comprehensive expression of robotic sensing, far beyond a basic grip.
It's like trying to pick strawberries with boxing gloves; the simple ability to grasp does not guarantee delicate and effective manipulation without crushing the fruit.