Let me repeat something, the electricity costs USD no matter what. If electricity is our largest operating cost, it doesn't matter if it costs thousands of dollars or dozens of Bitcoins... if I can cut that the whole way down to 1/5th of it, we save thousands, or we save dozens. That operating cost doesn't go away because we calculated the math in a different currency.
BTC prices since its inception have been rather random. Without an actual economy to back Bitcoin, it will continue to be unstable. Your example of BTC doubling is just as possible as BTC halving. We pay electricity in dollars, hell, we pay every cost in dollars, it simply doesn't matter what the price of BTC is now or what it will be in the future, we must pay those costs.
The point is that for a bit-coin mining operation (what Diablo claims to be) your revenues are produced in BTC not US$. Each block mined produces the same number of coins irrespective of the exchange rate.
When the exchange rate changes, the cost of mining (the electricity element) also changes.
If, at the current exchange rate, the cost in electricity to generate 1 BTC is X BTC (X would be less than 1) then were the exchange rate to double (twice as many US$ per BTC) then the new cost in electricty to generate 1 BTC would be X/2 BTC. This is a simplfication - ignoring the impact rising difficulty has.
YES - you still DO save money by having cheaper electricity. But it's less of a saving. If you're saving less per unit mined the time needed to reach break-even obviously also increases.
Further unless mining is inherently unprofitable (i.e. even with free electricty you couldn't make a profit) then there MUST be an elecricity price (expressed in BTC) where it's more profitable to invest in more mining gear than to invest in cheaper electricty costs.
This is indisputable - just consider the extreme where electricty were free (or as near free as made no difference). Hopefully you can see that at point there'd be zero value investing in cheaper power. Now there's also the other extreme - if the electrical cost of generating 1 BTC WAS 1 BTC then there'd be no way to make a profit UNLESS you invested in (or otherwise obtained) cheaper power.
Somewhere in between is the point at which investing in cheaper electricty makes more sense than investing in additional hardware and buying the higher electricty rate.
Now here's where you're going wrong:
1. Nowhere in this thread have you made any effort to calculate where that cross-over point is - yet the only possible financial justification for investing in generating power rather than using the same cash to invest in more mining gear is that there's some non-trivial chance that it's actually more profitable to do.
2. You don't seem able to grasp that the exchange-rate has a huge impact on where that crossover point is.
Let me explain point 2 for you.
Let's say we'd bought our hardware and were mining. We'd worked out that the point at which investing in cheaper power was better than buying more mining hardware was X BTC (X<1) to mine 1 BTC. Your power currently costs X*1.5 BTC to mine 1 BTC - so all looks rosy. But now the exchange rate doubles. Suddenly the cost to mine 1 BTC has dropped to X*0.75 BTC - and we're in territory where we're now making less profit (or more loss) than if we'd just spent all the capital on mining gear.
And if the exchange rate rises further then each rise just makes even worse the loss of value gained.
All I'm trying to prove to you here is that IF you denominate the investment in BTC then exchange-rate fluctuations can have a huge impact on profitability. I can't see why anyone would invest when you've produced no calculations or figures that address this at all.
There's even the theoretical possibility that exchange-rate fluctuation could make investing in solar-power completely unprofitable (expreseed in BTC) even if it ran for the whole 30 years life-span and all power was sold off at market rate.
You're selling shares denominated in BTC - so the base-point for calculation of profit is "how well would investing do, compared to leaving the BTC sitting in my wallet". Whenever you invest in something that generates product valued in US$ then the answer will ALWAYS be "that depends very much on how the exchange-rate fluctuates."
And when you're considering investing in two different things (mining and generating power) it's not a bad idea to work out how profitable each of them is - then you may well see you could make way more money (and have less complication) just doing twice as much of one of them with the same capital.
I've only referred to mining here - not to running a data-centre. That's because this apaprently is (or was supposed to be) a mining company.