But then you have ring species such as some seaguls. The gulls in Northern Europe can interbreed with the ones in Canada, which can interbreed with the ones in Alaska, which can interbreed with the ones in Siberia, which cannot interbreed with the ones is northern Europe. So is there one species or two, and if two where do you draw the line?

I think the point was: how can you know that when you only have a few kg of bone fragments?

Taking that as the metric how have the writers of the paper in the OP proved that the species that produced these bone fragments was not able to produce [fertile] offspring with any other species (of which they have a few bone fragments), or even with modern humans.

Even with a full genome, or a full skeleton is it even possible? Surely the only way we currently have to determine mating compatibility is to have live subjects to mate/extract gametes from and means to incubate the zygote [and care for it to adulthood].