There is some real concern that using range as the basis for a subspecies assignment, especially in species as mobile as birds leads to a feedback loop.
This is discussed here as well as several other threads on the forum.
Range can be used properly to assume a subspecies identity, I’d even go so far as to accept a statement that it can be used to assume with a high degree of probability. But the only way to ‘know’ that it is the particular subspecies is to be able to visualize and validate the morphological or other separator keys that define it.