No, the user put in GPS info, but it included an accuracy value of 104.86 km. iNat doesn’t show observations on maps when they have large accuracy values, but the exact cutoff value depends on latitude:
Taxa can be manually added to checklists, or they can be auto-added by iNat. That’s the buggy part though – sometimes iNat fails to detect qualifying observations, so they never get added. And if it does get auto-added, but then the observation is deleted or made casual, or otherwise no longer qualifies, iNat won’t remove the taxon from the list. So I might guess that someone posted a zoo-housed cheetah in the US, and it went to RG and then got auto-added to the list before it got marked captive.
The system to detect qualifying observations is too resource-intensive to maintain, and staff actually said at one point that they would remove it, but it’s still limping along.
A user must have set it to absent. I don’t think the system was ever designed to have observations override a manually set absent occurrence status.