>assuming that microsoft is a "typical" subset of the entire US economy As you note yourself, this is where the argument likely falls apart. Given that you ave

>assuming that microsoft is a "typical" subset of the entire US economy As you note yourself, this is where the argument likely falls apart. Given that you average wirh energy-intensive industries like aluminum production (well: any kind of production most likely) the figured might be not even in the right order of magnitude. Energy consumption is probably to a large part data centers (usually built close to energy sources like dams, though I'm not sure in the MS case), production and shipping of SKUs (less and less) and offices for 100k+ employees. Perhaps a large impact is also due to commutes. Are there any more tangible calculations around this?