Interesting video, I can't recall seeing similar. It looks how I feel in the mornings...
My advice is to remove the gimbal and inspect for any damage or dislodge components. Just a normal inspection after any crash... Any axis in a bind, motor or arm, could display like shown.
Understand the gimbal is in low-torque mode when it is simply on, high-torque when Solo is armed. Just making the point. As well if you flip Solo motors down the gimbal will enter "turtle" mode, basically cutting power to the gimbal motors.
You could try a level calibration. This procedure may help to correct.
I'm not aware of any access to the gimbal software. Others have mentioned using Mission Planner correcting odd gimbal behavior, but I have no first hand knowledge if it is a legit solution. Search the forum for further advice on using MP to correct.
The camera len's looks busted. Quite an impact or just an un-lucky tap? Sorry you had a crash, we've all been there once or twice...
That would be a "stick calibration", which has no affect to the gimbal. A "Level Calibration" is for Solo itself and does affect the gimbal, again search the forum for guidance as related to the "gimbal".
The USB mentioned elsewhere, is to access directly the PixHawk FC. This does not give you access to the gimbal.
After chechink for bent arms, I would check also for the roll motor tempereature. Even if it is just a little bit warmer than the others, it may be the culprit, if the axis are free to rotate otherwise. I had an eBay bought one with the same behaviour... interestingly on the same axis ... Turned out it has some windings in short, verified on an inductance meter (ohmeter may help, but very difficult to measeure the low resistance winding differences). In short, the motor phase can't deliver the required torque and can act like this.