I love that I can export RGB-averaged orthoimages. It's very effective at noise reduction using a small sensor camera (basically image stacking).
But I would be able to do more error analysis and supervised classification if I had more output options. For example, if I could calculate and export max/min intensity and standard deviation of the RGB pixel values, I could better identify areas of specular reflection.
Also if I could export standard deviation of the height map (or DSM), rather than the elevation value, for a given grid size, I could better analyze noise and accuracy. (and probably also better identify areas of specular reflection).
I could see this function implemented either as a calculator type function (maybe under "advanced" export options) or as a dropdown menu. Even if it was only exposed in the python API it would be very very useful for me.