Thanks for your answer. We are in another time zone, so apologies for the late reply. And yes, the build uses the two identical 185RPM motors.
On our previous build, the Robotic Arm Robot, the robot felt a little sluggish - there was a lag between events and actions (but it was of course controlled remotely over bluetooth). If this has persisted, it could be that when the robot gains some momentum going back and forth, the sensor data from the gyro may be processed too slowly for the motors to compensate in time.
When I hold the robot in the air, however, and tilt it back and forth on the gyro's x-axis, the motors spin slower towards the center of gravity and faster the more it tilts, before the motors stop at certain positive and negative values (when the program presumes it has tipped over). This all seems fine, which is why I am unsure how to proceed.
Do you have any suggestions as to what we can test?