Add ability to overwrite existing datasets in the `to_hdf5` methods of the various parsers

[wtbarnes]

In the to_hdf5 methods of the various parsers, in most cases, if the dataset already exists, nothing happens, i.e. the code will not attempt to overwrite the dataset. This was intentional to both avoid overwriting datasets as well as avoid errors if datasets already exist.

However, in some cases, it may be desirable to overwrite particular datasets or groups, e.g. #405.

For the parsers for the ion-specific files, this is easy because the group (and subsequent datasets) will not exist yet unless you want to override them. There are, however, some other non-ion sources which write data into pre-existing groups, e.g. the abundance files. These will require a bit more thinking.