Interesting thread! Without going into much detail (and there's infinite detail unfortunately!):
- There's technically no such thing as a 'species' in biology - it's a human term with an arbitrary meaning (and there's even multiple meanings!). In reality, biological variation exists across a spectrum, with gene flow capable of occurring between 'species' which we think of as being quite different (e.g., Polar and Brown Bears). If we take the most accepted definition, then our splitting/lumping should be entirely based off of reproductive isolation, where species can't (or can but almost always choose not to) interbreed. - Based on this, it doesn't matter how small the differences in morphology, vocalisation or behaviour IF these differences have resulted in a long-established reproductive isolation. - The problem this creates is, as you have alluded to, we then often rely on DNA analysis or tiny biometric differences which are often impossible to obtain in the field. But if we want to be accurate, these annoyances for us birders are actually not relevant in the splitting/lumping argument.
As a further aside - as you have said, we are currently in the midst of the 6th mass extinction, and splitting similar species often represents perhaps the best way of conserving isolated sub-populations with unique genetic lineages that would otherwise be lost if they remained a part of a larger, widespread species. For a non-birdy example, see the push to split the Giraffe into multiple species to aid its conservation.
Birding has for some time become an activity which relies increasingly on photographic evidence (which is fine) but also on the need to carry a specimen container to obtain faecial DNA in order to ascertain an identification in certain circumstances.
As I commented recently on Thailand Birding opinions remain divided on the identification of a Reed Bunting species (Common or Pallass) where the official decision seems to be coming down for Common Reed Bunting based on calls ( or a call) recorded despite the photos seeming to favour Pallass. Bearing in mind that Reed Bunting occupies a huge range across the Palearctic and displays noticeable variation in bill size - is it not to be expected that calls/songs will also vary, often with the possibility of overlap with a sibling species where ranges overlap?
What I am getting at is that we would never dream of splitting humans on the basis of vocalisation so why do we so readily split closely related bird species as if it is ALWAYS the ultimate clincher?
A couple of days ago I posted photos from Paul in Thailand of a Bluetail which remains unidentified. He now has drawn my attention to yet another possibility,- that of it being a Qilian Bluetail. This is a newly split species with a breeding range in mountain forest of northern China - winter range unknown!
Apparently this new species is identical to Red-flanked but has a different song - this being the (sole?) justification for splitting it. I suggest that it is inevitable that the song would be different if one population of Red-flanked lives in Finland and the other population lives in/around Inner Mongolia.
In conclusion - separation over time leads to change/variation by natural selection but are we splitting bird species prematurely (based on minuscule measurable variations) in a world where ironically we are in the initial phase of mass extinctions?
__________________
Challenges are inevitable, but failure is optional.
