I am contemplating creating a scientific journal. How many editors should I try to get on board?
Apart from the points made by Stephan Kolassa, I would enter the number of articles to be processed as a variable in the equation.
If you start a journal you probably are aiming for a certain number of articles to be put through review per year. Note that the number of articles processed can be substantially higher than finally published since not every submitted article becomes accepted. Rejections rates vary but 50% is not unheard of and higher rates are not uncommon.
So you have a certain number of articles to handle. I then assume that your editors are working for free outside their normal work hours. You have to consider how much work load you want to put on these persons. In the journal I "run" I have set a goal of 4-5 articles per year for each editor and have tried to adapt the editorship to that number. I am not saying 4-5 is a norm but you have to judge what is reasonable. You can ask potential editors what they can consider. Obviously they need to put some effort in so one paper per year is perhaps not optimal. The work load I consider concerns finding reviewers, getting the reviews back and recommending reject/revisions/accept (and completing additional rounds of reviews) to a Chief Editor for final decision (the latter decision process can be done in different ways and reflects "my" journal).
The scope of your journal also plays a role in that with a wide scope you may need editors with very differing specialities. You then have to balance the number of reviews for each sub-topic so that editors will handle the flow without overworking any one editor. With a narrow scope, all editors can probably handle all manuscripts and the problem is then "just" to distribute work fairly.
You can expect there to be difficulties achieving a balance from the start and assessing the likelihood of attracting papers and maybe how many, although largely a guessing game, will be important. This is where the answer becomes closely related to the question of how to start a journal because many considerations involved in making sure people publish affects the editorship of the journal.
This will depend on your field and on how specific you want your journal to be. If you concentrate on one very narrow specialty, you will be able to work with very few editors. If you want to cover multiple specialties, you will need different editors with different areas of expertise so each submission can be assigned to an expert in the field.
You may be able to learn more by browsing the editorial boards of journals in your field, and noting whether anything is said about specialties of the different editors.