A range of opinions were expressed on this subject in this thread recently (and in another thread linked a few replies down): ID based on observer notes
Personally I don’t think it’s consistent with iNaturalist’s standards of evidence to accept comments as sufficient to confirm an observation as RG. Otherwise there are plenty of observations which have no media which should be eligible for RG if they have a sufficiently detailed description of what was observed. But in pure text it’s more difficult to tell how experienced or honest the observer is. What if they misinterpreted the features in a dissection, or misread the graph indicating the chemical results?
For bat identification people will upload a spectrogram of the bat calls; I think here uploading an image of the results of the spectrometry would be analogous?