i would use this: https://jumear.github.io/stirfry/iNatAPIv1_observations?user_id=jimpark67&per_page=200&options=idextra&idextra_user_id=jimpark67.
here’s an example of the results that it returns, highlighting a couple of kinds of cases that might be of interest in this case:
in the first kind of case, you have an observation at subspecific rank, and the identification of the observer in this case is at an ancestor taxon to the observation taxon (species or higher).
in the second kind of case, you have an observation at species rank, and the observer’s identification matches the observation taxon. however, there is at least one ID for a descendant taxon. this means that someone other than the observer has added a subspecific ID.
if you can define what kind of cases you’re looking for, then it is possible to write a script that will get all your observations and then on the client side filter your observations for only those that match your criteria. in my mind, it would probably take more effort to do this than just to visually scan through the results from the page above.
this post sort of describes how i might generally approach that sort of custom filtering, although it doesn’t directly explain how to handle this specific case: https://forum.inaturalist.org/t/insight-in-disagreed-annotations/55181/8.