That is a problem that also comes up when converting a boat from Bermudan to junk rig. I recommend joining the Junk Rig Association (
junkrigassociation.org), where you find the largest number people who have either converted boats to junk rig, or designed from scratch.
One question they are likely to ask is which kind of junk sail you intend to use. A cambered junk sail is likely to have its centre of effort a bit further forward than a flat sail. A split junk has more area before the mast, so a centre of effort closer to the mast, so the mast can be further back. The split junk doesn't allow much adjustment of the balance area, the others allow you do adjust how much of the sail is before the mast by where you attach the halyard to the yard, and by using parrels.
You could set up the boat with a board in front of the centre of effort, and some lateral area aft, either a long keel, or a large rudder. (If you use a rudder, it helps if it is balanced.) Then you can adjust the centre of lateral resistance. When you sail on the wind, drop the forward board to get what degree of weather helm you desire. When you sail off the wind, raise the board, giving the boat as much lee helm as you want. That may give you inherent self steering over a good range of courses to the wind. I once saw a 10 metre long shallow draft junk schooner in the harbour, and spoke with the owners and designers. Mudskipper had three boards. Two in the bilges, far enough aft that lowering only one or two of them gave lee helm, and one forward, which could give weather helm when lowered. While the owners were designing Mudskipper, they tow tested various models in a river. Their main criterion was control. They did not want a boat that would broach. The multiple centreboards were the solution, along with a canting rudder.
Length to beam ratio also matters, though perhaps it is more hull length to mast height ratio. Mast height tends to be proportional to beam. So, for a given heel angle, the centre of effort moves out to lee by a distance proportion to rig height. So if you have a short hull relative to rig height (and beam), then your rudder has a short lever arm to counter the weather helm that you get. I had a Benford Dory 28, and it developed very strong weather helm when pressed hard with full sail. I am sure that had nothing to do with the junk rig, and everything to do with the boat being rather short for its beam, with a modestly sized rudder. In a 15 knots wind with full sail and perhaps 30 degrees heel (I think a boat should be under control at that heel angle), I needed 30 degrees rudder, which seems excessive, and was probably slower than if I had reefed a panel. But I wanted to see how the boat can handle being pressed.
Your design seems to have similar length to beam ratio as that Benford Dory, so I would recommend either a large, balanced rudder, or that forward board so you can adjust the centre of effort to where you want it. My dory had a fixed keel only.
Phil Bolger said that the first time he designed a sharpie that had the same curve in the bottom and side panels (having no curve in the side panel is, I think, specifically a trait of New Haven sharpies), he did so only to make the boat easier to build. To his surprise, he found the boat was easier to control downwind, and generally faster. He came to think that is because equal curvature in the edges gives equal curvature to the panels, and so equal pressure either side of the chine, and therefore fewer eddies across a sharp chine. I never came across an explanation how that works once you have leeway. Perhaps having eddies of the same intensity all along the chine helps? Whatever the explanation, Bolger says the boats sail better. The drawback was a stem that is out of the water, giving a flat overhang that can lead to slapping at anchor. And with the flare you get from the angle of the side panels, you would get a lot more rocker, so a lower prismatic coefficient. That tends to be an advantage at low speeds, a drawback at higher speeds.
Also, Bolger says if you can round the chine, that can also prevent those eddies. The Dovekie (I think about 5.5 metres or 18 ft long) has no rocker at all, but a 10cm (4 inch) radius chine takes care of the eddies, at least with the curvature of the side panels it has and at the speeds it can reach. So if you want to go with Bolger on this, you could put in a chunky chine log that you shape to be round on the outside, or you strip plank just the chine.
If you have a river nearby, do tow tests with a model. Set up two masts, and a longitudinal spar in between to which you can tie your tow line, so you can move the tow line fore and aft. Attach the other end of the tow line to a pole. Give the model a radio control, then steer it so that the tow line pulls at angles ranging from directly forward to as close to abeam as the model can handle. Do so with different heights of the longitudinal spar over deck to simulate reefing, and with different attachment points along the spar to simulate different locations of the centre of effort. If you can find a longitudinal placement that lets you control the boat at all attachment heights and all reasonable speeds through the water, that is where you want the centre of effort to be. If you need to move the towline fore and aft to balance at different flow speeds, different courses, and different heights, then your hull would benefit from an adjustable centre of effort.
Also, that hull does look good.
