i’ve seen observation fields used in ways that could be classifying observations under undescribed species. for example, here are some cases where @megachile seems to have highlighted some gall observations as a possible undescribed species: https://www.inaturalist.org/observations?verifiable=any&place_id=any&field:Similar%20observation%20set=dryocosmus_q_phellos.
the problem with observation field and traditional project approaches is that not all folks allow others to add fields to their observations or include their observations in projects.
you should be able to add an identification in most cases though (unless you’ve been blocked by an observer), and if you know that you can identify things to a specific placeholder taxon – the genus or a similar species – then you could just place comments/notes in the particular identification and then look for those identifications with those specific comments later. the main difficulty with this approach is that you probably need some familiarity with coding and the API to efficiently extract a list of observations marked this way.