Regarding automation of the interface between HF and a microscope:
1. On my Olympus stereomicroscope, I added a stage to support the object and a dial indicator with one inch range reading in 0.001 inch increments easily resolved to follow the head of the microscope. For some pix I move the head 0.001 inch at a time at 35X and as much as 0.005 inch at a time for low magnifications. My Olympus camera can be operated with its internal self-timer to expose the image as few as three seconds after I press the shutter release. The exposures range from one second to 1/125 second, depending on the object. This gives nice results, and most of my time is spent waiting for those three-second delays to expire. It's all very flexible, and the object stays in place just fine.
2. On my metallograph everything is just about the same (same camera, for example) except that I use the calibration of the fine focus knob to space the stacked layers. At high magnifications, the increments have to be in fractional microns (0.001mm) but the results look like color scanning electron microscope images at similar magnifications. Same three second delays between exposures, but sometimes a great many images, say 200. HF never chokes on this quantity of images. My light source is a one watt white LED. I'm limited only by the working distance of the objective lens.
The first Olympus camera logged more than 20,000 images using HF. I'm now using another Olympus camera (E-620) because it has Live View, which saves a lot of guesswork about where to start the stack.