Sea,
The angle at which the gyro is tilted while resting on the mains is not arbitrary. It's the angle at which the nose and the tail both become "weightless."
All pusher gyros of which I'm aware have the main gear axle behind the CG with fuel and pilot aboard, and so it is with the machine Paul tested. This allows us to taxi with the nosewheel on the ground. To make it balance on the mains, you'll need to raise the main gear on blocks and tip the machine back until it just balances, meaning you could let go of the keel at both nose and tail ends, and it would just hang there.
As with the plumb line through the teeter bolt in the hang test, in this test, a vertical plumb line passing through the main gear axle will also intersect the machine's CG. It will do so at an angle different enough to produce an easily seen point of intersection with the one from the hang test. That intersection is the machine's CG.
Remember that there are no separate "vertical CG" or "longitudinal CG." The CG is a single point, and we're concerned here with its relationship to the thrust line, usually considered to be the same line as the prop shaft. In most fixed-wing aircraft, the datum is treated as a line or plane to simplify calculations of balance, because the CG's vertical location is never an issue, but the actual CG is a point.
There is also a "lateral" location by which we could describe the CG, but most aircraft are so close to symmetrical left-to-right that the lateral location of the CG can be assumed to be in the center when the aircraft is viewed from the front or rear.