Dan, You have to remember that a Miller Table is not designed for processing mass quantities of materials. It's a "Finishing tool" and as such it's limited by several factors. Size alone is going to limit how much you can run at a time. Raw materials will require a longer length to give the gold time to settle to the bottom so it can be caught.
I feel that you would be much better off with a sluice that is about 3-4ft long X 10 inches wide and a good pump that puts out at least 1500 gallons per hour. If you up the pump to 2200 gph you could run the sluice flatter and have a better chance of catching more of that micro gold. Angle, flow, feed rate and surface are all factors in gold recovery. Keep it level side to side, adjust the flow so it's not blowing the materials through the sluice like a jet engine and trying different drop angles along the length and with a little trial and error you'll be able to recover 90%+ of any gold in your materials. Once you've concentrated the materials down with the real sluice, run the cons on the Miller Table. Depending on the materials you could use several different types of mats to catch the gold. I'd recommend using a couple different types in the box. Having the same stuff all the way down the sluice is insanity. Different surfaces (ribbed mat, moss, expanded metal etc) will each catch gold in different ways. If you only have a single type of surface in the sluice it only gives you one way to catch the gold. Read up on changing recovery surfaces on the GoldHog site. Doc does a much better job of explaining it than I can.
EDIT: Getting every bit of gold is not an easy thing to do. Even the Big Boys loose some during processing. How much you get will depend on how much time you want to spend on it. Some folks (like me) save up black sands after we've gotten all the visible gold out to process further. Things like crushing, roasting or amalgamating take time and many don't want to dedicate a small lab to be able to do these processes in a timely manner. The extra gold you get out of these processes isn't much, but I'm an old lab rat and like to do this kind of stuff.