The H33D CAN have the camera installed upside down but you have to create a custom mini-profile USB plug that will be just the plug and nothing else. Nearly have to have bare wires for it to fit. I will do this. Like I said it might make BOTH landing gear show! Or none... Moving the entire gimbal forward is possible as well but needs to be accompanied by a similar rearward move of something else of substantive weight on board. I will explain why this balance is critical in my view shortly and this could explain a number of the issues I am hearing on these forums. One way to balance the Center of Gravity again if you move the gimbal forward is to move the battery connector inside the P2 BACK by the amount needed to balance the CG change rather than to add weight... Add a shim that is exactly the width and profile size of the existing back mating plate on the P2 so the battery will still clip in. The shell is styrene so you can use a piece of thick styrene and recreate the little lips that the battery clips attach to. This way the battery will connect be held firmly in place while providing enough mass aft of the CG to allow for the forward move of the H33D gimbal (or any gimbal).
Personally moving the phantom battery connector back by the 4-5mm it might need may work just fine...
Now even if you DONT move the battery back slightly to compensate for a forward move of the entire gimbal system I am sure the P2 will fly fine if the camera mount is moved forward a cm or two . I really do think so.
Here is the important element here I think: The problem is that moving weight around and shifting the Center of Gravity on any multicopter will cause some of the motors to work harder to maintain level balanced flight. This wont be noticeable in general and you can take off and say "Hey that works great!" but it will cause definite dangerous changes to turn and descent profiles and more importantly allow the Vortex Ring State onset to occur easier. The reason is one of physics... If you have one or two motors, say the two forward motors, working HARDER to maintain level balanced descent for instance, if a gust of wind hits the P2, it is likely that the quad can be more easily destabilized by the sudden change in the forces. During balanced flight there is no appreciable net difference between the upward thrust vectors within a narrow acceptable range I would guess. If that upward thrust vector is offset by a gust of wind hitting the hanging camera system, then the forward motors, overthrusting in the case illustrated here will tend to force the front end up at a FASTER rate than if the forward and rear motors were all at balanced thrust. This will tend to destabilize the craft easier. Its best if all four motors feel equal balance forces. I see people putting all kinds of neat gear on the P2 and making side trays and all... What they dont realize is that although the NAZA can certainly compensate for this, the additional strain on those motors to maintain level balance will allow failure conditions easier as outlined above.
My suggestion is that any changes to the system should be accompanied by a careful re-balance or re-weighting to maintain similar thrust vector values. Its theoretically possible that the increase in runtime gained by preventing overthrusting even in the simple case of level hovering flight would counter the loss created by adding a tiny bit of CG balance weight aft for instance. Run time might even improve for all we know at this point once the load on the motors was more balanced. Just my two cents based on a completely scratch built 52" diameter quad prototype that we built where we had to tune PID loops (the Proportional Integral Derivative setup from the Ardupilot platform). We found that even with the hanging weight balance setup (center of gravity a little bit below the plane of the propellers) that out of balance conditions would cause issues that the autopilot could not counter.
Another suggestion for stability when making changes is to get the motor arm extenders I have seen online. Adding two inches or so to the diameter of the P2 would enhance stability, increase the rotational moment and make it a little harder for VRS to grab hold. But, THAT wreaks havoc with carrying cases!!! haha...
So these are just my thoughts if it is helpful to anyone.
Thanks,
Marc Dantonio