I have some experience with this approach. All of my reefs are fishless, as this is the only way to eliminate bioload and equipment hassle in pico reefs (less than 3g).
It is very true that feeding a system provides nearly all essential nutrients in terms of protein and vitamin availability for the hardier corals. Stocking selection will affect selective nutrient uptake (such as the notable Ia fixation in xenia) and mineral depletion (stony coral calcium/alk demand). Copepod numbers are easily maintained in a non-predated and well-fed system, after all what else are fish doing for us (nutrient wise) other than processing the food we add. You would be suprised what a healthy copepod population can consume, this is sufficient to keep suspended particles/compounds available for hardy reef organisms.
Were it not for copepods and reef insects, this feeding regimen would not work in a nano and it would soon be ridden with bacteria and algae. Non-reduced food particles are substrate for bacteria, which will increase oxygen demand in a nano threefold--> quickly smothering everything aerobic. Copepods compete for detrital matter, keeping bacteria and even oxygen demand in check in the balanced nano reef.
I like to use a variety of scavenging inverts as well, so I can still have motion in the fishless aquascape. These guys require food input like the fish, and excrete similar metabolic compounds that can benefit coral in small amounts. The boxer crab Lybia Tessellata is the hardiest crab and nearly the best colored Ive ever seen, one small juvenile survived a temp spike that killed every coral in my system. Banded coral shrimp are nice I think.
My theory is that we don't have to keep fish with corals to have an invert-focused tank. The compounds supplied by fish presence are also generated by organisms on roughly equal trophic levels, so the key to a CO (coral-only) system is not adding corals before copepod populations are up and running.
Brandon M.