The short answer is that IDs are a community process on iNaturalist, and different identifiers will have different levels of confidence based on multiple factors of experience and quality of the evidence.
The general expectations are that each identifier will identify as specifically as they are reasonably confident doing so. That may vary from Kingdom level to subspecies level, depending on each individual’s experience with the organism and/or the geographic location of the observation, on what evidence they are able to see in the observation, and on what additional information the observer may have provided in descriptions, comments, or their own ID. The community ID algorithms behind the scenes sort out those differences to arrive at the “official” consensus at any given time.
If you haven’t already, it’s worth reviewing the Identifications section in the FAQs also.