Overhaul of dark matter PrimaryFlux class and data

[StFroese]

The recent issue from @micheledoro #5147 shows the need for a more flexible class for the PrimaryFlux class, which allows different spectra files for input.
Currently there are three different options (maybe more):

1. PPPC4DM file that is available in gammapy-data already
2. HDMSpectra, which is package that provides newer spectra https://github.com/nickrodd/HDMSpectra/tree/master
3. CosmiXs, which provides PPPC4DM-style files https://github.com/ajueid/CosmiXs/tree/main

This brings me to the question if we should provide these two other options in gammapy in addition to the first one and if yes, should the other spectra also be provided in gammapy-data?

Be aware that the handling of the keys for the spectra table might be different, it could be the case that it's not easy plug and play with the spectra file.

Additionally, one might be interested in not only the flux for gammas but maybe also for other primary particles like neutrinos if we look ahead at a possible combination of instruments from all of the VODF experiments.

[micheledoro]

Some more information/comment:

• a student collaborating with me had prepared a parser from HDMSpectra format to PPPC format. However, we did not do all the complete conversion process. And I am not sure it's needed, considering CosmiXs state their work is more accurate
• I agree about the CosmiXs file added in gammapy-data. Although I now always prefer to set the location explicitely in the notebook and not hidden in an automatic way
• about neutrino flux, I am not sure gammapy is the right way to handle from inside at the moment, but if the file contains nu-spectra then it's straightforward to have them extracted no?