Split up documentation in actual model part and reference implementation part

[johanneshiry]

Currently, the documentation is heavily entangled with regard to the actual bare model description and the reference implementation. As more and more people make use of the actual model scheme, but not necessarily the java reference implementation, we should split up the documentation parts into the actual model documentation (incl. things like assumptions and handling of null/empty values, assumed folder structure for file based data etc.) and the documentation of the java reference implementation. Having this would make the understanding easier and avoid confusion.