I’m not a gall researcher, but this is what I’d suggest. Unless you open the gall up, you don’t know which life stage the organism is currently at (could potentially be pupa, larva or maybe egg) or even if the organism which induced the gall is still there (it could have been consumed by a parasitoid, for example). And even if you did open up the gall, the larva or pupa in it could be a parasitoid.
All you know is that there is a gall, which is recent evidence of the organism that induced the gall.
Good point. A somewhat weak retort would be that the person wasn’t adding metadata to the observation, they were adding the observation to a project, which is different. But I don’t think that’s a particularly great distinction. ;)
This motivated me to do some spot checking of observations for some of the projects and observation fields you listed because we really don’t know. In my non-scientific analysis, there were more observations that would not meet the Gall annotation definition in projects than with the observation fields. But neither were perfect. Which kind of calls into question the entire idea of retrospectively auto annotating obsevations, IMO.