From The Book of WoW: Rule 1837b: Never look forward to an effect that depends on a proc, thou shalt surely be gnashing thy teeth and will soon be in the greatest dudgeon with all men who walk the earth, forsooth, for great delusions will tumble about in your phantasy about coding and other matters which lie beyond your understanding.
I think you guys went abit off topic here :)
Did some statistics on numbers from raids with nightfall. Cant link the excell file but can summerize it here. I did tests as a 2h fury dps with nightfall and calculated the dps from myself plus the added casters dps to give my "true dps". I did avg dps of 368, our casters did average dps of 351 and the dual wield fury warrs did an average dps of 578. This means a loss of 210 dps for me. Thus did nightfall have to compensate for that 210 dps by increasing the casters with more than that amount. The average uptime of nightfall that raid was 16,5% which added an average of 8,94 dps per caster. It whould in that case require, at 16,5% uptime, a total of 24 casters in the raid to be worth it. I played around with the model and tweaked some values. Our raid team had seven casters which whould require nightfall to have an uptime of around 60% for it to be worth it. To summerize; same conclusion, its &*!@. Function for required uptime (y) depending on amount of casters (x): f(x)=136e^(-0.0955)x