I guess part of the problem is, there may be several solutions for developing the surface between arbitrary chine, keel and stem curve, but not all of them produce a desirable shape, and we are only interested in those that the plywood will naturally choose. As it stands now we need to situate the controlling splines and the controls of those splines in such a way that the modeling software finds the reasonable development solution easily.

I think that there is only one solution that is truly developable in the sense that there is zero Gaussian curvature and no twist along ruling lines (and no hard creases). I guess the word developable is often used in a less strict sense, and as soon as the requirement is relaxed and compound curvature or twist is allowed, you get an infinite number of solutions. Picking a good one is what I am struggling with, as I cannot find any design criteria for the allowable twist. What really surprised me is that the perfectly developable solution for this problem appears to be the worst one, i.e., it generates the smallest bending radii. It is easy to compute bending strains for thin plates given the curvature and estimate whether plywood can handle the curvature from available material properties. The twist that is introduced along ruling lines is much less intuitive and the only design criteria I can find in the literature are for metals (of non-specified thickness).

My current thinking is that 4mm plywood must be able to handle quite a lot of twist. At least, I have not found a way to model a surface that appears to be buildable without introducing significant twist in ruling lines. Some of this may be a side effect of how I generate the designs. The maximum twist appears to be quite sensitive to the transition from keel to stems, and this is somewhat tedious for me to analyze. Like you said, adjust - check, adjust - check, only this takes quite a bit of time with the tools I am using. Your previous post about trimming surfaces has given me an idea I want to explore. Maybe I can write some code to automatically adjust the stem to reduce twist.