3 on each side (control and experimental) is certainly better than 1 on each side. Of course, 6 is better than 3. My thought, if you have a particular budget, space allotment etc...I would work towards fewer varieties of corals, and more individuals. Thankfully, there are corals that are often readily available in large quantities to frag (i.e. Monti Cap...get a big sheet or two from someone and you can easily make a bunch of dime sized frags to test). LPS is hard...softies is vague. I could see this test being much more relevant if you stick to a few choice SPS that are rock solid and inexpensively available...afterall think of having a frag rack with 6 digitatas, 6 caps, 6 seriatporas, 6 pocilloporas, and maybe 2 X 6 acros. That's still only 36 small frags in each 10 gallon tank...not a lot of corals in the grand scheme of things, but since these all have similar requirements you can light appropriately, maintain water chemistry, reduce or rule out aleopathic issues, and focus really on getting good sized data samples from multiples of similar types of corals.