Why use the uncorrected angular velocity in observation model

If the GPS is mounted at an offset from the CoG, any angular velocity from the vehicle will induce a velocity due to the rotation, which is unwanted and needs to be corrected for.

In the GPS measurement model, this is accounted for, by subtracting the induced velocity caused by the positional offset. However, the uncorrected raw angular velocity is used.

What is the reason to use the uncorrected angular velocity?