Potentially misleading meaning of `positionError` in `Cluster`

[tmadlener]

Triggered mainly by this dicussion.

[jmcarcell]

Comparing to other types, we have covMatrix for TrackerHit3D, and TrackerHitPlane for the covariance matrix of the position. We also use the same name in other places

EDM4hep/edm4hep.yaml

Line 608 in 5968932

- edm4hep::CovMatrix4f covMatrix [GeV^2] // covariance matrix of the reconstructed particle 4vector

and

EDM4hep/edm4hep.yaml

Line 189 in 5968932

- edm4hep::CovMatrix6f covMatrix // covariance matrix of the track parameters.

, so the most consistent option would be to rename to covMatrix. Then there is the question if we would like to specify what these covariance matrices are for in their name.

[tmadlener]

I think the Cluster is a bit special in this regard, because it has a position and an error on that, but it also has a physical extent / an (intrinsic) direction. So there is two quantities that can have a covariance matrix attached to it. Whereas in e.g. the TrackerHits it's somewhat obvious that the covMatrix refers to the position, similar for the TrackState.

Arguably a similar argument holds for the ReconstructedParticle. However, the fact that it has a 4D covariance matrix might be confusing for other reasons (see #423).