I do this sometimes. Usually, it’s when I’m not sure if my phone’s microphone has picked up the sound well, so I also mimic it so that it’s clear what the faint, far off noise really sounded like. In some cases, my phone’s microphone is not good enough to record what my ears heard and so my mimicked noise is all I get on the recording.
Rarely, the noise has stopped by the time I get my phone app open. In that case, I think it can be useful to mimic what was just heard while it’s fresh.
I don’t do this often, and I don’t expect iNat will ever be swamped with such observations. Still, I can see it being useful occasionally. We certainly do need to be clear when the recording does not contain vocalizations of the animal being observed (eg marking “no” for “Evidence of organism” on the iNat website).
This discussion reminds of that video of a cockatiel whistling the Andy Griffith theme song. So a recording could be ambiguous – is it evidence of a cockatiel, or only of a human (on TV)?
i sort of don’t understand why this is such a controversial subject. a drawing is just another way to record an image. an imitation of a sound is just another way to record a sound. sure, these methods introduce some error, but even cameras and audio recorders aren’t necessarily capturing things exactly as they are. (for example sometimes people take blurry photos or record muffled audio, or sometimes encoding algorithms introduce artifacts in the images or sounds.) we just have to trust that folks are using the best tools and best efforts available to capture their observations.
folks are free to ignore these kinds of observations, and certainly nobody has to suggest or confirm an ID if they don’t think the evidence supports it.
personally, i have nothing against text descriptions as evidence and i think it’s always a nice complement to photos and audio, but the reality is that the overseers of the system have decided that that kind of thing is not adequate as evidence for the purposes of an independent identification, and i can understand why.
the difference between drawings and vocal reproductions vs a text description is that the way the text is decoded by human brains is highly variable. not everyone is going to be able to decode a highly technical description properly because they may not share the same vocabulary and definitions. but a less technical description will be interpreted vastly differently by different people.