It can be thought of as the inverse of the "Heightfield from bitmap" command. That command takes an image and creates a surface that has its highest point at the lightest value of the image (like white) and its lowest point at the darkest value (like black). Intermediate values in the image (grays) will have intermediate heights on the surface.
ShowZBuffer does the inverse - it makes a black and white image from the different "heights" of (all) the geometry in the viewport (as seen from the camera perspective). The lowest point (furthest away from the camera) is black, the highest point (closest to the camera) is white and other values are various shades of gray...
Beware: the grayscale calculations are based on the viewpoint of the camera and (I think) the automatic clipping planes Rhino uses to define the furthest and nearest points... sometimes it can get confused.
If you want to use this image, you need to export it with ViewportCaptureToFile or ViewportCaptureToClipboard (the second requiring pasting into an image processing program). And if you don't want something to be included in the height calculation, you need to hide it first. --ch