Hello mpatalberta,
camera_location_accuracy parameter defines the precision of the measurement device, so this value is fixed and wouldn't change due to the processing workflow.
If you need to get the difference between the estimated camera locations and the source values, input to the Reference pane, then you can use the following code:
import Metashape
chunk = Metashape.app.document.chunk #active chunk
for camera in chunk.cameras:
if camera.transform: #checking only aligned cameras
estimated_coord = chunk.crs.project(chunk.transform.matrix.mulp(camera.center)) #estimated XYZ in coordinate system units
source_coord = camera.refence.location #source values in the Referenece pane
error = (chunk.crs.unproject(estimated_coord) - chunk.crs.unproject(source_coord)).norm()
print(camera.label, error)
This code should print to the Console the errors for the camera locations as difference between source and estimated values.
[/code]