When I was first designing Phoenix III, a major consideration was proper rowing geometry. Rowing is a wonderful way of providing auxilliary power, but it is very important to get the rowing geometry correct - done well and the rowing will be a pleasure, but if it is wrong the rowing will be an unpleasant effort.
There are a number of considerations, but the two most important are: - 1. The horizontal distance from the aft edge of the rowing thwart to the oarlocks, and 2. The vertical distance of the oarlocks above the rowing thwart.
Because these are relatively small boats, there is not a lot of vertical distance available under the rowing thwart, which limits the width (or chord) of the centreboard. Not only does this mean that board area is is reduced, but it also means that the "bury"of the board in the case is very small when the board is lowered. This means that the board and the case are heavily stressed by the side loads when sailing.
My solution to the problem was to make the board and case wider at the forward end - something which used to be common in working boats such as Swampscott Dories. The end result is a board which has much more area than the rectangular one, and it has a lot more "bury" in the case when lowered. Also, the open part of the top of the case is higher in the boat, which helps keep things dry.
The only real disadvantage with this system is that it makes it impossible to have a thwart for a forward rowing station, as the oars would strike the centreboard case. The forward rowing station is important if you want the boat to trim properly when rowing with a passenger on the aft seat (stern sheets), but I judged that Phoenix III and First Mate were primarily sailing boats, and adequate centreboard area is very important for good sailing performance to windward.
However, I've received quite a few requests from plans buyers who really want a forward thwart, and so I've started investigating the possibility of putting in a rectangular board. It will have to be wider, and therefore the rowing thwarts will have to be higher. This in turn means that the oarlocks will have to be mounted higher. But I think it is workable.
As you can see, the rectangular board is still a little smaller in area, but I might be able to fiddle around with the length to make up the difference.
Regarding the underwater shape of the original board, I ran a spline through the apexes of the angular outline, and you can see how it approximates the shape of a shark's pectoral fin. Probably a good structural/performance compromise.