Well, sorry for the probably false information following. Those are just what I understand of the process, and you shouldnt take this for reliable information.


Anyway, I think there's a (actually, several, nested) control system (I mean, some closed loop control) dealing with this. So basically, the head follows the pregroove. If you send a fake ATIP signal to the drive unit, this control system may get fooled and think the head is still on track. But then, you have near no information on where the head actually is, or you have to bother for it by yourself.
This control system's role is to have the burner always write on the pregroove. If it thinks the head has lost track, it will raise a burning failure signal. So your mod system must manage to fool this control system... and basically, I dont know if the area outside the pregroove can physically be burnt...

On a normal CD, or when a burner reads, two systems can come into play:
- the lateral sensors & associated control system, which would 'see' the wobble OR the datatrack itself in case of a big offset of the head relative to the datatrack, and/or when jumping from one radial position to another
- the central system & associated control system (central relative to the head not meaning main system), responsible for beam focalization and angular speed adjustment. It keeps the beam focused but in normal conditions it may help to stay on track. Anyhow, my guess is that there's some PLL system recovering the clock : the pits are indeed organised as multiples of a spatial period, which makes a given temporal period when read at a given angular speed. The reading unit has to have a correct angular speed to see correct pit lengths.


Remember that all I said is just my small theory about how this works, and that I cannot assure you any piece of that is true. If anybody has precise facts, I'd be really pleased to see my thought corrected