Here are some possible sources of error:

1. Manufacturer stated focal lengths are often rounded favorably or sometimes plain exaggerated (e.g. a 200mm might well be only 190mm).

2. Focal lengths are also stated at infinity focus, lenses change focal length when focusing (though it is more common for a lens to get wider as you focus closer not narrower).

3. Sensor sizes are not always specified accurately (if this is how the AOV was calculated), even within the same format from the same manufacturer there will be variations of around 0.1mm that effect your angle of view.

4. RAW editors crop your images, as most demosaicing algorithms use neighbourhood operations (where the value of a pixel depends on surrounding pixels) in the edges of the image there are no neighbouring pixels so these areas are cropped out of the final image. Usually the amount lost is only around 4 pixels from each edge.

5. Lens distortion corrections applied by the camera JPEG engine or during RAW conversion change the effective focal length and/or result in the image needing to be cropped.

6. Measurement errors in your 120m or 216m figures.

Some or all of these adding up are enough to explain the discrepancy. If lens corrections are being applied that will be the biggest factor and could probably explain the result on it's own. The next most likely/largest source of error will be in the 12mm focal length figure.

Here's a worked example. If the lens focal length is really 12.4mm and the sensor is 0.1mm smaller than spec then your diagonal angle of view will be 79.61°. If you're really at 117m and if lens corrections + demosaicing are causing you to lose 3% off each edge then you would expect to see 183m, which with less than 2% error could be measured as 180m.