updates to Sr2RuO4_SOC

[phibeck]

Update to the Sr2RuO4 benchmark:

• replacing the tight_binding_model() function by that in lattice.utils
• SOC lambda matrix could in principle be used from dft_tools/converters/wannier90, but different orbital order so it would end up not being much shorter
• the notebook contains results from an actual calculation, but there aren't any results in the directory, so perhaps one should update the notebook?
• we can do this calculation in cthyb, so we should add results