You can define the zero point to be wherever you want. Only changes in potential energy are relevant, not absolute values. So, if you define the GPE to be 0 when the car is on a level road, you're defining the GPE to be 0 at a height of 0.1 m above the road. So really the GPE in this case is given by mgy, where y = 0 at the position 0.1 m above the road. A height h above the road corresponds to a position of the centre of mass of y = h-0.1 m above the point where the GPE has been defined to be zero. Hence the GPE here is mgy = mg(h-0.1).
Instead, you could have defined GPE to be 0 at h = 0, in which case the car would have GPE mg(0.1 m) on a level road, and a GPE of mgh at the point specified in the problem. But this does not change the fact that the difference in GPE between the two points is mg(h-0.1)
An analogy is elevation. Really only differences in elevation matter. I could measure elevation from sea level, in which case, if I was at sea level, my elevation would be 0, and if I was 300 m above sea level, my elevation would be 300 m.
Alternatively I could measure elevation from the centre of the earth, in which case my elevation at seal level would be approx 6400 km, and my elevation at a point 300 m above sea level would be approx 6400.3 km. But the absolute elevation values don't really matter so much as the difference between them. Nothing has changed in the second case. The higher elevation point is still 300 m above the sea level point.
