That being said, I noticed something peculiar when checking out a document cited in the Wikipedia article. The SDK says 0.4 is the “about the perpendicular drag of a cylinder.” If they are using the typical drag coefficient, 0.4 is about equal to the drag coefficient of hemisphere and sphere, but not a cylinder.
The circular cylinder is more like between .75 and 1.2 (see Fluid-Dynamic Drag : Sighard F. Hoerner : Free Download, Borrow, and Streaming : Internet Archive page 3-16, figure 28). The drag coef. never goes below about .7, even for a cylinder with an aspect ratio of 1 (e.g. 1" span x 1" diameter). Logic would dictate the coefficient would get bigger the more squared the cylinder is, and lower the more elongated its cross-section in relation to the direction of airflow. So perhaps fuselage_lateral_cx should be sitting more in the range of like 0.7 to 1.3 rather than 0.2 to 0.8.
It almost makes you wonder if the folks writing the SDK saw the picture of the sphere and assumed it was a cylinder. A sphere has much less drag than a cylinder. Since I have little knowledge on the subject and no engineering background, it’s pretty much speculation on my part. But if there’s something to it, that would still be pretty significant for developers to know.