Interesting… to extend it further, the footprint of an animal is a reproduction, not the actual foot.
It’s getting into semantics more than it needs to be though. I think it comes down to how reliable the evidence is, and how critical the pin location and date accuracy are. If we are arguing that the accuracy is critical, then goodbye all the feather and shell obs… but anyone analysing the data is going to factor that sort of thing into their work. Would anyone working with range data on a given bird species be factoring in the likelihood that the observations might be of mimicry and present a good distance from where the actual live bird was when the mimicry was learnt? Highly unlikely. But for shells it is unrealistic to think of them not considering it!
Also, in the case of mimicry of calls, I think it would come down to a species level thing… how likely is it for that species to remember and use the call throughout it’s life, or to only use the call when the mimicked bird is present or near by? I think there are too many unknowns to be able to treat it reliably and consistantly