It shows how to construct the pose estimator, but not common usage like how to update it with encoders (like the odometry class) or how to apply vision measurements.

We should also document how to get those vision measurements in the first place, because that's not trivial. PhotonVision has utility functions and docs for that we may want to consider replicating.

https://docs.photonvision.org/en/latest/docs/programming/photonlib/util-calculations.html