I would like to start my comments with a disclaimer. I don't have a well known reputation in the stunt community i.e. no name flier however I talk to a great many people and absorb information well. I spoke on this subject to a great many people not the least of whom was Pat Johnston who has designed and built many planes. This topic was of great importance to me because I am in the middle of designing and building a semi scale stunter based on the bf 109. The low wing placement on the prototype leads to problems with the vertical cg in flight. I have decided to add dihedral to the wings in order to deal correctly with this problem. The real issue with the vertical cg placement is that as stated before, if the weight mass is above the leadout position, the airplane will fly wing low while upright, suggestions indicate that by using a tab or adjustable "aileron" you can correct this. This is true for upright flight. In a perfect flight where there is NO speed variation this may work. however this tab adjustment will be by design speed sensitive and therefore introduce other trim issues as our planes do speed up and slow down throughout the pattern. The correct way to offset this is to get the leadout position in the correct relationship to the vertical cg, or for understanding, the center of mass. To check for this relationship in an assembled airframe that is ready to fly, hang the airplane from the leadouts in a door jamb, orIEnt the airframe so that it is "flying" through the door, ie from one room to the other. Assuming that your door jamb is truly plumb the distance from the wingtip to the jamb ideally should be the same on both wintips. To understand this phenomenon and how it affects flight, imagine a bucket of water on the end of your lines. If you hold it by the handle it will swing straight out in line with your lines or arm. If you hold it by the edge of the bucket, it will instead find its own center so that the mass is evenly divided above and below the lines. This is the problem with vertical cg issues. As has been discussed before, the bellcrank does not care where it is, its all about where the leadouts are. Most of these discussions center around fore and aft issues. However the same is true in relationship to the vertical cg. If you have ever seen a "spirit of St. Louis" model fly control line, you will note that there is a hanger off the wingtip to locate the lines below the wing. This is a dramatic illustration of how to correct for vertical cg location. In my case, and Al Rabe, and Pat Johnston the solution is to raise the wingtip up to more closely approximate the center of mass in flight. Hence by using dihedral we move the wintip and more importantly the leadouts up into a more correct location. Al Rabe could more correctly describe it than I have perhaps but that is my understanding. By moving the engine vertically away from the designed and flight tested location, we have altered the vertical cg location and must somehow accomidate for this.The same holds true for rotating and engine that was designed inverted into an upright position or outward position ie as in a profile. Understandably there are more than one way to skin a cat but this is my take on it. I don't like using aerodynamic bandaids to fix a mechanical problem. If you would like to read more, there is some additional posts in the ARF section under the two posts regarding the Brodak P-40 and some things people have observed and dealt with regarding it. It was designed as a low wing with dihedral, however in production it became a low wing without dihedral and people are moving the wing, or adding dihedral to correct this issue.
I am open to any and all comments on my stance please feel free!