A collected (killed) specimen has not undergone further development after the date of collection – i.e., the life stage, phenology, etc. all represent the point in time of collection.
In this case, however, the adult flies were not present when the host was collected.
I don’t think the original poster was considering using the date of collection, the question was whether the location should be where it was collected or where the flies emerged.
For me, the guidance on iNat is clear – it should be the location where they were observed several months later, not where the specimen that carried them was collected, even if this is outside the expected range of the species. Data like this is always going to have to be analyzed in the context of the circumstances of the find; the point on the map doesn’t tell the whole story. Linking to an observation of the host at the time and location of collection seems like a reasonable way to represent this history.
Some of the discussions on raising caterpillars and hatching gall wasps may be relevant (though these count as “captive” at the time of emergence because the organism being observed was intentionally collected rather than an accidental passenger). See e.g.:
https://forum.inaturalist.org/t/seeking-guidelines-for-hatched-insect-galls-on-plants/21752
https://forum.inaturalist.org/t/captive-raised-caterpillar-observations/31361