That's almost a peak oil approach, which was both insufficient and the most accurate possible. We can only state things in terms of stuff which currently exist

That's almost a peak oil approach, which was both insufficient and the most accurate possible. We can only state things in terms of stuff which currently exists. The future will yield different results because new techniques and technologies will come along. But we can't rely on or predict them. There's a lot of "pons and fleischmanning" when talking about future energy use - mass scale geo-engineering or sucking carbon out of the atmosphere technology, it has to be avoided. The technologies don't exist yet. We have to use real-world stuff. Using current reality for future projections is inherently defective because of the innovation factor and routinely leads to bad predictions (such as peak oil[1]), but it's the only reality we have and is sadly as good as we're going to get. Anything else (such as say next-gen nuclear) has to be regarded as a fiction in the calculations until it actually exists. It's messy, sloppy, and requires a large trashcan but the alternative is to get to results that rely on things that literally do not actually exist. Kinda like how the people planning Operation Downfall (the land invasion of japan during ww2) weren't told about all the other things in the works (backroom deal to get Russia to declare war on Japan and the Manhattan project), they had to assume those things would continue to not happen. [1] - The Hubbert peak was essentially that the amount of energy to extract and refine oil would surpass the energy from the oil and thus make it unsustainable so the operations for oil extraction would cease. New techniques have been developed (such as fracking) which have changed this math, but in 1956, using 1956 knowledge, this was the best prediction possible. It's somewhat Malthusian - who was also as "correct as possible" with the best science that 1798 had to offer. He couldn't factor in, say, the Haber–Bosch process from 112 years in the future in his food growth computations.