Refactor authors similarly to peers

[vphantom]

Big can of worms, but we should handle authors the same way we do peers, with full manual acceptance of the copyright contract, e-mail nagging for this acceptance, and a new status in the workflow which prevents moving on until all authors agreed.

This implies locking the list of authors after that point, to avoid the impossibility of having an article under review or worse, with a missing copyright contract.