First of all, it's important to note that Google's Sycamore has not (yet) processed any quantum algorithm, but instead has just proved that the quantum processor is capable of maintain quantum effects on all 53 qubits during operations. This is important because there's still controversy on if all quantum computers built to date exhibit real quantum advantages.
In order for this processor to become a "practical" quantum computer, also measurement physical errors must be lowered.
Also, to make computations that can actually break current cryptography algorithms, the number of qubits must probably be raised, to around 1000 or more, all depending on how low the error rate will be.
And answering the second question, no, this result is in line with prospections on quantum computing, so the (optimistic) estimations of around as early as 2027-2040 to break ECDS still maintain. But all depends on current trends that nobody knows if can be maintained or not, as this is all based on cutting-edge engineering and mathematical advances.
Also note the link indicated by @ugam-kamat regarding practical mathematical implementation of cryptography breaking algorithms that also points to a minimum of around a thousand usable qubits to break nowadays signatures. This would also raise the real number of needed qubits by some factor n>1 (n>>1?), as the previous noted increment was due to measurements errors which would still apply as the number of qubits raises.