I forgot to mention: The Younger Dryas temperature change was not completely global: in particular Antarctica would have missed it.
But the CO2 change WAS global thanks to excellent mixing. So a CO2 signal from ice from Antarctica shouldn’t be uncritically compared against an Antarctica temperature proxy. Compare the CO2 behaviour against as many global temperature proxies as possible.
It's obvious that there are a wealth of global temp proxies in a cloud around the CO2 results. The Shakun authors must have used an industrial cherry-picking machine to find the proxies that suited their purpose.