I am using the longterm average because thats how I read data. The more info the better. We could cut everything out pre asic, but those increases wont affect the #s that much, so there is no need.
Again, from a statistical standpoint, this parabolic rise can not continue. I am not saying it will stop tomorrow, but it will have to slow eventually.
Look at it like this. Assuming 10phs where we are equals 1.5billion in difficulty, to increase the difficulty to 3billion, IE 100% increase will require 10phs more of hardware coming online, if all the chips were cointerras 500ghs chips, 10,000 chips would need to come online. And this doesnt account for all of the older gens of hardware that are being made obsolete with every increase, nor the hardware that is likely dieing or being pointed at altcoins.
In what way does this alter my conclusion?
Its info for your brain to chew on.
My main point being 2 fold I guess. 1) Why were you basing your math on a site that is claiming a 1% increase when its doubled in the past 3 months? 2) The "compensation agreement", isn't that important for a couple of reasons, a) they are on schedule, and b) Its the reinvest that matters to keep with difficulty.
1 last thing.