....
My point was simple. Your calculation is based off of a static total hashing value for ActM. That value will not be static for any mining operation. It will grow with the TNH to maintain the target percentage. Because you assumed a static hashing value for ActM your whole calculation, while already being hypothetical, is even further off.
You are whining over an unimportant detail. Complicating estimate calculations will not make it more correct and is usually total waste of time. BTW, maybe you should look up what word "random" means.
Having a goal to maintain a percentage form TNH on it's own, is flat out stupid. Problem is, at some point it makes no sense to add new hardware at cost X, because this hardware will not earn back the investment made. This is like buying GPU cards or outdated FPGA's today and getting all exited about your rigs total hash rate, while you can never earn back your additional investment - total waste of money, time and energy.
How fast can Actm add new hardware to it's mining operations makes sense only to certain point in time and difficulty. After that point, it is useless to add new hardware, unless it's more efficient than chips Actm going to gave in November.
What ever they paid per chips, assembly and maintenance of running rigs, has to be calculated in and it becomes permanent part of the cost X.
Selling portion of your inventory will not lower the cost X.
I hope Ken understands this and sells as many rigs as he can. Those first
lucky smart buyers will make assload of money with this equipment. Ones who wait, till it's too late, will be left holding their dicks. Sure, they can move over to PPC or similar alternatives. (Hmm... can PPC be mined with ASIC's?)