Add tooling to get more revealing data from the sensor

[svanbergaddi]

We need more tools to be able to produce more reliable IMU data

Gyroscope

Requirement to move to in review:
Data being returned that enables next part of process to calculate median and peak velocity (a whole another layer)

[svanbergaddi]

Three-part pipeline: subscribe to Gyro alongside Acc → fuse them into an orientation estimate → rotate acceleration into world-frame and integrate to velocity, with drift correction (raw integration drifts without correction — the standard trick is "zero-velocity update," resetting velocity to 0 whenever the bar is detected at rest between reps).