The wayback machine!!! As a programmer it feels like you've just gotten an ice cream off my grandad!
There are much more complete versions of those graphs I've seen around, Garf linked me to them once, I'll see if I can find them. They hint at food input being more than enough to negate trace element dosing if water changes are scarce.
Lastly, the word "preliminary" has been mentioned, that means at least 50 pages on this thread!