A common question with drone imagery is accuracy specifications, something that is pretty established and understood with satellite providers, but highly variable with drone flights.
Does it make sense to add spatial accuracy estimates for drone imagery as a(n) (optional) metadatum? I think most of the major COTS sUAS post-processing software provides some measure of this, and OpenDroneMap will estimate this too.
Related:
#6