There's no perfect CG point since the final CG point is related as much to the comfort zone of the pilot as the design of the airplane.

Having said that and provided you do not have any serious angular setup oddities in your model I would suggest that Lou hit the nail on the head with his description. The CD is an aerobatic model and such models should not need to have more than one or at MOST two degrees of downthrust compared to the wing airfoil's centerline. If you have issues with ballooning at that point then they ARE CG related.

By adding downthrust to excess you're actually masking the true culprit which is the CG issue. The best way to test this is with some dive testing. Set your throttle for about 1/4 to 1/3 of the max RPM (just a little above idle in fact) and trim for level flight with the elevator trim. Now push the model into a 45 degree dive and let go of the stick and watch the model. If it has a very abrupt nose up reaction then your CG is too far forward, if it has a noticable but not fast nose up then it's likely just right for general sportflying, If it stays in the dive or wants to tuck the nose harder then the CG is too far back. For something aerobatic like a CD the altitude needed to pull out of the dive should be large. A slow pullout that uses 250 to 300 feet is not untypical. Often pilots will set the CG and resulting elevator trim so the model doesn't pull out or tuck at all. At dead neutral in other words.

The CG descriptions seem to be at odds with the results but they aren't. In the first case (and what it sounds like you have) the CG is strongly forward (I don't care what the plans say). A forward CG is more stable but it requires more up trim to compensate. In a dive where the speed builds up the elevator trim will become stronger than the excess nose heaviness and pull up the nose sharply. Similarly when you add throttle the speed builds and the strong elevator up trim needed to hold up the heavy nose at slow speeds becomes too much and again the nose comes up.

To get a proper handle on the issue you need to measure all the angles of the model's setup to ensure that something is not out of whack. This includes the engine thrust angle ( measure the prop disc and take away 90 degrees), wing airfoil center line angle and stabilizer center line. Set all controls to dead neutral to blend the airfoils perfectly before you do this. The results may find a problem that you need to deal with. You're looking to have the stabilizer at 0 to -1 degree to the wing and teh engine at 0 to -1 or -2 at most compared to the wing. From there you just need to work on that CG.

Yes your landing charactaristics will alter a bit but it won't be that much. You may have to slightly change how you approach and set the model down but you should still be able to land it fine.

By the way, part of your "problem" is the engine. You've overpowered it and that will also tend to make this ballooning issue that much more noticable. WIth the extra power the need for properly setting the angles closer to 0-0-0 and the CG towards a more neutral setting becomes that much more important.