Therefore, you can measure acceleration with nothing but equipment (a simple spring-scale will suffice) in your car (reference frame), but you can only measure your speed relative to something outside of your car (traditionally, the road).
Newton's first law (inertia): an object in motion tends to stay in motion. This gives rise to the idea of an "inertial reference frame" which is the idealized reference of a hypothetical object in motion. In simplistic terms, if you pretend that the Universe is merely Newtonian in nature and contains a preferred reference frame then you could imagine that an inertial reference frame can be described by a velocity vector (a speed and a direction). If an object has zero speed relative to the inertial reference frame that will tend to continue to be true so long as no forces intervene.
If you induce an acceleration on an object you will change it's velocity. In it's initial inertial reference frame it will appear to be accelerating. In any other inertial reference frame it will also appear to be accelerating.
There is no such thing as an accelerating inertial reference frame precisely because such a frame is not inertial, it would imply that forces are being applied to the objects involved. The laws of physics make sense if you apply them within any inertial reference frame, but they don't make sense in accelerating reference frames (mass, energy, and forces become inconsistent and confused).
Does that make sense?
Similarly, rotation rate isn't relative. Acceleration and rotation rate can both be measured against absolute states of no acceleration and no rotation (respectively). This is why our phones have accelerometers and gyroscopes which are entirely internal, but determining velocity, position, and direction require the external references provided by GPS and compass.