Document DLPS124C, page 16, figure 6-14 appears to show that test pad B17 is solidly within, and pads A17,A18,B8,B18, and C8 are partially within the Thermal Interface Area, where a presumably conductive heat sink will be in contact with the device via a rectangular hole in the PCB.
Question 1: Is this a misreading of the document?
Question 2: Are these test pads able to be safely shorted to the heat sink, and thus presumably also to the heat slug on the back of the DMD?
Question 3: Is it recommended that the heat sink be milled to avoid contact with these test pads, which would naturally make it have less contact area with the DMD?