Now you are simply posting incorrect or misleading information. There is no existing implementation of Grover's algorithm on quantum hardware that demonstrated any speedup over classical computers. In fact the largest implementation of Grover's algorithm to date, on ibmqx5, has an unusable error rate.

> Literally the only difference between the state of the art and quantum supremacy is scale.

This is simply incorrect. At the moment it's unclear that quantum supremacy can be achieved at any scale whatsoever, and it's certainly possible that it's unachievable even in principle.