The more I think about it, the more I'm thinking that getting the C.G. right is going to be crucial for transitional attitudes and airspeeds between a balanced hover and forward flight. This is my thought process:
longitudinal (fore-aft) C.G. in a balanced hover isn't really going to matter much. It's going to have some effect on the stability of the aircraft, but that really doesn't matter since we're using a flight controller to stabilize it anyways. The elevons also don't need to be super effective in a balanced hover, because as long as the aircraft stays vertical, the forces the elevons need to produce to counteract any small attitude excursions are relatively small. But when the aircraft starts to pitch forward and the wing begins to develop some amount of lift, all that changes. Now the wing is producing lift, presumably behind the C.G, causing the nose to pitch further forward towards a forward flight attitude. The further the nose pitches, the further behind the C.G. the center of pressure moves, and the tendency to pitch forward continues. At some point the wing will be producing more forward pitching moment than the elevons can counteract, and the only way to recover would be to increase the elevon effectiveness by increasing airspeed, power (I think this is what Leonard's patch is trying to do?), or both. A good example of what is happening is this: In slow, low power flight, if you hold full up elevator on your aircraft, and let it stall, does the nose go up or down? On anything I've ever flown, it goes down, and that's exactly what is happening here.
But let's look at that situation with the C.G. moved back to a neutrally-stable location. As the airplane pitches forward and the wing begins to produce lift, it's acting more or less through the C.G and it shouldn't cause any pitching moment. The elevons should still have the same effectiveness, and they should be able to bring the nose back to a vertical position. In forward flight I don't want to fly something that's truly neutrally stable in pitch, as it can be a real handful, but since I'm flying with a flight controller anyways I can abandon our aerodynamic stability and replace it with electronic stability in FBWA and higher flight modes. As a bonus, I should gain a little bit of efficiency in forward flight. I think this is part of the reason why the tailsitter code flies @tridge's addictionX so well... it's designed to be pretty neutral in pitch stability, and has huge pitch control authority so it can do elevators, ect.
I think that before I apply leonard's patch I'm going to try shifting the C.G. back 1 or 2 cm to see what the effect is.