I just found out that there is an identification method on iNaturalist in which only the first photo is shown and comments are not shown. This method involves just identifying off the panel in the Identify tool without actually going into it. (It has been here a long time probably, but I never noticed the “Agree” tabs right on the panel.) This makes me think that there should be a separate DQA category for multiple-species observations to immediately take them out of “Needs ID.” (Same for duplicate observations.)
That explains a problem I’ve had repeatedly had where I submit an observation with my ID explained in the comments, usually citing a recent paper on the species, and then others disagree with the ID as if they hadn’t even read the comment.