First of all, one of the tracking principles of a SteamVR is that the tracking algorithm needs to see at least four sensor locations in the form of a tetrahedron. A tetrahedron is three points in one plane (say the X-Y plane) and a fourth point at some distance from the X-Y plane (Z). So you want to have sensor locations on the tracked object at varying heights to create the maximum number of tetrahedrons possible on the tracked object. One way to do this is to 'sink' the sensor location within the plastic shell of the tracked object. Once that sensor has been 'sunk' beneath the surface of the plastic shell you need a way to get light down to the sensor. That's the purpose of the concave apertures you see on tracked objects. These concave surfaces maximize the angles at which the swept laser light can hit a particular sensor.