Simplifcations and fixes to the two-scale heat conduction case

[IshaanDesai]

This PR does the following:

1. Removes the data K01 and K10 from the coupling because it is zero. The data is zero because the off-diagonal terms of the upscaled conductivity matrix are zero. This is due to the microstructure's symmetry (circular grains). If, in the future, a non-symmetric microstructure is designed (ellipsoid grains), the data K01 and K10 will need to be added back, as they will no longer be zero.
2. Update the configuration visualization as the number of coupling data is reduced.
3. Several fixes in the micro-nutils participant.
4. Add docstrings to functions in micro-nutils/micro.py to make them understandable.
5. Change direct-access flag in precice-config.xml to the new api-access flag.
6. Turn on lazy initialization of the Micro Manager to speed up the tutorial. Lazy initialization means that a micro problem is created only the first time it is active. This speeds up initialization because micro problem objects for inactive problems are not created.

Checklist

• I added a summary of any user-facing changes (compared to the last release) in the changelog-entries/<PRnumber>.md.

[MakisH]

All very welcome changes! Since you are now cleaning up, two minor points.