First, the author info is by definition (or rather by policy) identical to the user info in my journal and in every forestry and natural resource journal I have been publishing in during about 30 years of my experience with paper and electronic submissions. What is different in yours? (the data consists of: First name, Middle name, Last name, Affiliation, Country, Email, URL, and Bio statement) Beside, that’s irrelevant anyway because such selections would be optional and editable, while every author is a potential reviewer.
Next, the editor DOES have the access to the entire list of users already, and he uses it to select the reviewers – right? So, what is so hard to make the same kind of function for the author selection? When it comes to editor’s correcting of the authorship, which authors cannot do after submission, he (me) ends up going to the registered users in one browser and coping and pasting the info from it - bit by bit - into the author editing window in another browser. This is very frustrating and inefficient.
However, this time I appreciate your negative-spirit reply – it gave me a new great idea: there should be (in the same place) also a link to making an account for any author who is not registered yet. You’ve got the bulk of the needed data for it right there all ready to use – one click and you save a bunch of typing with no typos.
Finally, about the “presupposes” – if you use the email as identifier you’ve got it guaranteed.