Certainly, the architecture of a robotic arm like the Festo Bionic Handling Assistant may seem to improve manipulation, but claiming it intrinsically reduces the fragmentation of robotic vision systems is a hasty generalization. Where are the observable metrics of this 'reduction in fragmentation'? How do we concretely measure the integration of sensors and vision without a defined threshold? A parcel sorting system that fails against a damaged box due to a software failure is a clear example that mechanical flexibility alone is not enough.