J.C. is correct, of course. A gyro uses only a fraction of its longitudinal cyclic control range in flight. Bensen-style gyros have a very large (30 degrees aft) tilt limit. This limit is seldom, if ever, reached in flight, and not usually even in landing. Bensen provided this large tilt to assist with starting the rotor in the absence of a powerful (or ANY!) prerotator. Air & Space and McCulloch gyros, both with full-RPM prerotators, have less available aft tilt, I believe, without creating any control problems.
Someone should ask Jim Vanek if he pulls the stick all the way back to the stop to execute his loops. I doubt it.
Whether you use elevators, a swashplate or a tilting spindle, in the end you still change the rotor's orbit via cyclic pitch change. The interesting things about elevator pitch control in gyros are that (1) the airframe LEADS the rotor in changing its angle of attack, rather than following it, and (2) elevators work equally well in all G-loading situations, while direct-tilt systems lose their power to control the airframe at low G. The second issue is the root of the problems with low-G control of gyros that have direct cyclic control and centered flapping hinges. One function of the H-stab on a direct-cyclic gyro is to make sure the frame follows a stable flight path during time intervals when there is little or no pitch control, thanks to low/zero G.
PPO is an extreme example of the frame NOT following a stable flight path during low/zero G.
The great weakness of elevator control is that it loses effectiveness at low airspeeds. Immersing the elevator in the prop blast helps, as long as the engine is running. Control will still be poor during engine-out flight, though.