Assuming good will, the researcher may want to compare (1) the initial automated identification, (2) the community-sourced identification, and (3) the identification derived from molecular techniques/expert taxonomists. That would seem an interesting study!
I would hope, if that were the case, that they would acknowledge the work put in by identifiers on the site, and even offer the possibility of collaboration and co-authorship on an eventual publication to the top identifier(s). That would also have the benefit of including an author who understands well how iNaturalist works. It might be worth reaching out to them, assuming good intentions, and suggesting such a collaboration.