Move Wyckoff to positions

[ndaelman-hu]

So, the Wyckoff positions are high-symmetry points in the conventional unit cell. This should be clearly stated in the descriptions. This is important, as not every crystal structure is reported in this structure. In the NOMAD front-end, we do, in fact, report the conventional unit cell in the topology hierarchy whenever possible.

Any Wyckoff analysis provides 2 kinds of information:

1. Generic tabulation of the positions of the high-symmetry points. This is fully determined by the space group (type of symmetry for crystals), and the setting.
2. Labeling of which atomic positions correspond to which symmetry positions.

I'm unsure whether we'll provide any search functionalities for Wyckoff, but it is super-useful for inferring the effects of a site on all kinds of observables.

[ndaelman-hu]

In the NOMAD front-end, we do, in fact, report the conventional unit cell in the topology hierarchy whenever possible.

We could already do this in the data section, adding a new model_system in conventional units. This may be useful for other kinds of analyses down the line.

I would have the Wyckoff sites mimic the atomic positions, and add further detail about each type of site to a repeating subsection.
MatID uses the Wyckoff sites and atomic number to assign unique locations too. Given that the information (and more, like atomic settings) is all at the same level, I would remove this quantity.