Bit of a late reply but I think I understand the question.
Colourized depth data is in the format of a depthmap. A 2D image whose pixel intensities represent distance from the sensor. The colourized version converts the depth intensity at each pixel from a grayscale value to a more visually friendly RGB value. The colourized version is to allow a person to better visualize the depth.
A depthmap can be converted to a point cloud if the user puts in a little work. The depthmap provides the Z value but the user needs to calculate the X and Y value. Using the pixel location on the depthmap, the focal length of the sensor and a little bit of trigonometry, you can calculate the xyz position of each pixel on the depthmap.