I must point out that there are two issues here, one is the immediate response of the helicopter to control movements (dynamic control response) and the other is the place the cyclic comes to be when the pilot has adjusted for all the aerodynamics of forward flight (static control trim.) Vorticey is noting the fact that the static control trimmed position is moving steadily out to the right as he goes faster and faster. This is due to several factors, not the least is the lateral balance of the helicopter. I suggest that vorticey make careful note of the difference between this trim for times when both seats are filled, and perhaps when he swaps seats while solo (is this allowed?)
We can make large changes in the static trim angle with lateral cg, and also when we trim the high tail rotor to more or less thrust in forward flight (the tail rotor tries to roll the aircraft, and lateral stick is needed to cancel that factor). Also, if we move the horizontal tail from the left to the right side (assuming it is only on one side of the aircraft) we get large roll cyclic trim shifts, as much as 15% of stick travel with everything else left unchanged.
The other posters have focused on the dynamic control response, which is controlled by the relationship between three different control reference frames:
1) the control input axis (forward cyclic stick)
2) the swashplate tilting axis (usually about 90 degrees before the control input axis)
3) the rotor tilt axis (where the tip path plane dips the lowest after a cyclic input)
The pilot might note that when he pushes an inch of forward cyclic, the tip path plane might not go purely forward, but might have some roll as well. This particular effect is due to many contributers, including gyroscopic precession, rotor flapping resonance frequency, rotor blade flapping inertia and flap aerodynamic damping.
The amount of difference between the control input axis and the swashplate tilt axis is called Gamma by designers, and is easy to change with no rotor head modifications. All you have to do is make the pilot's controls mix differently so that the desired tilt is caused. This is done in larger helos by a mixing unit (as in the S-76 and Black Hawk) or by a simple double swash plate (as in the Sea King). Both schemes simply make a forward cyclic tilt the swashplate in a non-forward direction of the designer's chosing. In flight test, we tune the cyclic controls (through adjustment of the variable mixers we have on test aircraft)until the forward stick truly achieves a forward tilt of the rotor plane.
It is surprising to many people to find that the Gamma angle is seldom exactly 90 degrees. If the designer guesses wrong before flight test, and is not too energetic after testing to make it right, you can have non-orthogonal control system that makes the pilot adjust for yet another helicopter control anomoly.
------------------