You can not explain the much higher GDP/oil barrel of Japan versus US because Japan manufactures less than US. Nor can you explain the declining oil intensity of OECD countries because they are importing manufactured goods from China. Even if all of China's oil consumption was included in G7 countries, they would still would have decreased oil intensity in last 30 years.
A lot of manufacturing doesn't use much oil, think of steel( coal), cement(coal) plastics( natural gas) aluminium( electricity). What uses oil? ; transportation especially cars and trucks, air transportation. Rail and ocean transport only uses a very small portion, and in most countries almost no oil is used for electricity production. Some oil is still used for heating homes and offices, but this use is being replaced by electricity or NG.
If oil prices continue to increase, its reasonable to assume continued improvements in GDP/oil barrel, in fact improvements in GDP/energy unit, although there is no reason to think that energy use will decline as oil use declines.
China doesn't have to reach the 12 barrels/capita, because they are building high density cities with mass-transit infrastructure.

Because in a service economy you get some other country to build your stuff. The West would use rather more energy if it had to make all its own t-shirts, CDs, cars and so on.

The IEA is probably forgetting that it still costs energy to make the stuff, where on the globe that barrel of oil is being burned or turned into plastic wishbones doesn't matter to resource depletion or climate change.

At some point the West has to start making things again. The job exportation to China and other parts of the world is not a sustainable situation. I don't see how you could borrow goods forever or how you can trade services internationally equivalent to goods. So any IEA analysis that anticipates the current international manufacturing distortion continuing or even increasing would be based on a false premise in my opinion.