Definitely the weight. Not only weight, but perhaps more distribution of the weight,
moment of inertia is more significant than CG, since the model is not just an effective
mass rotating about the CG point. It is being rotated in pitch, and being constrained to
something approaching a tangent to the flight hemisphere. And variables! Good God,
there are so many, everything from the airfoil, stab, flap, elevator, rudder and fuselage
side areas, wing area and volume, wing/flap/stab/elevator ratios, wing contour and
sweep, wing tip design, wiggling rudders, line diameter, gyro effects, drag, thrust,
lift, control displacements, engine torque/RPM power characteristics, engine cooling
and heating effects, fuel oil content and viscosity, prop efficiency, line drag,
density altitude, temperature, control system non-linearities, turbulation,
airflow lamination/separation, humidity, airspeed, groundspeed..
Oh yeah, and *COLOR SCHEME of the model! (What did I leave out?)
L.
*e.g. if you use a top dark/bottom light paint scheme on the fuselage with a
separation line which is not very close to horizontal in level flight, the
optical illusion can make the model appear askew in pitch..
"The rule is perfect: in all matters of opinion are adversaries are insane." -Mark Twain