Modern automotive audio systems increasingly rely on spatial processing, voice interaction and personalized listening experiences, placing new demands on in-cabin acoustic measurements. Traditional microphone- and array-based methods provide repeatable and standardized data but often fail to capture the acoustic influence of a real occupant. This presentation examines how in-cabin measurement methodologies are evolving toward more realistic, human-centric approaches. Building on established practices, we discuss the role of automotive-focused head and torso simulators, such as the new GRAS AutoKEMAR Head & Torso Simulator, in bridging objective measurements and perceived audio quality while maintaining repeatability and relevance in complex vehicle cabins.