It can not be a function of IRS because it is not related to acceleartin in any direction
I have no idea just how the various boxes process sideslip in the real world. However, the above statement certainly is not correct.
If the kit detects an acceleration, it can integrate that observation to figure a velocity. If the acceleration then stops, the velocity doesn't (in the absence of an additional, different acceleration). Knowing the side velocity and forward velocity vectors, it is a trivial calculation to figure slip angle.