Align Einstein Telescope geometry with LAL

[milanwils]

Problem
As mentioned in #713, the current geometry of the ET is not in line with its definition in LAL.

1. The labels of ET2 and ET3 are swapped.
2. Currently the new corner points are calculated by updating the latitude and longitude approximately and not updating the elevation. Additionally, the azimuth and tilt do not take into account the curvature of the earth. This leads to small differences in the vertex position and unit vector. As a result, the null stream is not accurate and signal leaks into it.

Proposed Solution

1. This is addressed in this PR by adding an optional clockwise option to the detector config. I set the value in the ET.interferometer file to clockwise=False so that the default is consistent with LAL. If this is undesired for backwards compatibility, it could be set to clockwise=True .
2. This issue is addressed by following the same specification as LAL: T1400308. The code is rather verbose to avoid confusion but can be changed if denser code is preferred. This algorithm required the implementation of a transform from WGS84 coordinated to ECEF coordinates. I added a small test to check the consistency with the forward transformation. In the implementation of this transform I used uninformative variable names to be consistent with the reference (they do not have a physical meaning anyway).

Additional information
There is also a small error in the LAL config, see LAL PR. Only if both get accepted do the configurations become identical.