As John K was alluding, AoA of -10 or -15 are not really practical and do not correspond to a powered dive like you think.
Even in a descent or a dive, most aircraft will have positive AoA.
You are likely confusing deck angle, angle of attack, and flight path angle.
Angle of attack (alpha) is the angle between the aircraft reference line and the velocity vector.
Flight path angle (gamma or theta) is the angle between the velocity vector and the horizon.
Deck angle is the angle between the aircraft reference line and the horizon. It is the sum of AoA and FPA.
In a descent or dive, you have a negative flight path angle. If your descent is steep enough, you may even have a negative deck angle. You likely still have a positive AoA.
In a steady climb or descent, L = W cos(theta). CL=L/(q*Sref). CL=f(AoA).
For negative reasonable angles, cos(theta) is positive and even close to 1.0. At negative 10deg, cos(theta)=0.98. In that 10deg descent, your CL is 98% of what it is in cruise! If your aircraft has symmetrical airfoils and no incidence (i.e. CL=0 at alpha=0), then your descent AoA is 98% of the cruise AoA. Most importantly, it certainly has the same sign.
It is sometimes counter-intuitive, but climbing, descending, and turning can be treated as 'steady' maneuvers. For climb and descent, steady means that the accelerations are zero.
Being in a steady climb/descent is separate from the dynamic process of initiating that climb or descent. That dynamic process is considered maneuvering flight -- unsteady flight.
It is possible to reach negative AoA during a maneuver in a way it is not possible during steady flight. So an aircraft may experience negative g's during the push-over to initiate the dive -- but in the steady dive, the aircraft experiences flight very similar to level flight.
Edits follow
It would seem unusual for a conventional aircraft to exhibit the characteristics you describe.
Sideslip should have small effect on pitching moment.
A stable aircraft should have a dCM/dCL that is negative -- which implies that at negative alpha it should have large positive CM.
A very quick VSPAERO run on a wing/tail combination shows this.

This shows CM vs. CL for AoA from -10 to +10 and Beta from -10 to +10.
Here is a view of the solution at AoA=-10 and Beta=10.

When building your VSPAERO solution, start simple and build to complex.
A better place to ask OpenVSP questions is on the OpenVSP Google Group.