@jsitarek would you like to review this? It comes from your old scripts linked here under "MAGIC PROPOSAL TOOLS / Source Detectability"

looks good, just one typo, but since you are either way using gammapy, why not taking the EBL models from it?

yes I am going to use gammapy in the end for most of the estimations, this PR is a bit legacy: I'd like to make sure that the package replicates the old scripts first before the big refactoring.