The IRS position, relative to the nominal centre of gravity position, is known. The IRS can determine it's own movement, both in terms of translation and rotation. So it would be entirely possible to correct IRS data to any other position on the aircraft - and this is done in some experimental aircraft instrumentation setups. The error from having sensors very far offset from the centre of mass can matter. Though conversely, since the pilot is also not at the centre of mass, and it can be a pilot sensation you are interested in, sometimes you want the sensors offset.
So it is true that the raw IRS data is affected by the position in the aircraft, but that doesn't necessarily apply to any processed data.
@Vessbot - yes, the pitot/static system is going to be affected but the effect here is less that the static ports are moving relative to the ground and more that the aerodynamics around the nose is going to be changing as the aircraft rotates, and this effect is very much dependent on exactly where you are in the ground effect, and won't be exactly the same as the PEs you might see for similar angles of attack away from the ground. Obviously all depending on the specific design.