It was always a long way away. It's a hardware problem, trying to implement some of the most delicate controls ever, at the coldest temperatures ever. Just enough to rotate qubits but not enough to decohere them. Then as you scale up, you run into more problems with correlated errors as qubits start interfering with each other. The algorithms have already been developed, for the most part. All the theorists are just waiting for the manufacturing to catch up. Probably another 10-20 years before you see serious industrial applications.