Currently it's only possible for the author to update the paper from the arXiv which isn't ideal - once a review is complete it's entirely reasonable for the editor to want to be able to pull the latest version from the arXiv with the changes that have been made as a result of the review.
Are we going to assume that auhor and submittor are the same thing?
Also logic wise, a paper is assigned to a specific editor, but should we say that any editor can take an editorial action.
Currently it's only possible for the author to update the paper from the arXiv which isn't ideal - once a review is complete it's entirely reasonable for the editor to want to be able to pull the latest version from the arXiv with the changes that have been made as a result of the review.