I’ve seen this argument pushed unironically, and quite convincingly.

It of course depends on a lot of factors, and GHG emissions are not the only concern, but “short-circuit” consumption can (apparently, I did not run the numbers myself and read this a few years ago) emit much more CO2 than importing food from far away… simply because driving a car for 10 km to a farm for a bag of apples (or whatever) is a LOT worse per apple than the traditional container-on-ship->container-on-rail->semi-truck->local store supply chain which has a few times the fuel consumption of a car… but multiple orders of magnitude more cargo.
This is in reality not so much a dig on short-circuit consumption, which is obviously overall good, than a dig at how polluting cars are, even compared to cargo ships whose emissions we intuitively over-estimate. Still, it has stuck with me as a good example of the complexity of making a life-cycle emissions assessment.

Modern globalized economies are also often criticized to have gone too far into economies of scale, making them very brittle… as we saw in 2020/2021, as farmers re-discover every time one illness destroys an entire country’s mono-culture, and as we fear we may discover soon with TSMC.
Furthermore almost every country (even very economically liberal ones like the US) heavily subsidizes their local agricultural sector to shield them from foreign competition, as it is of the utmost national security importance that a blockade on agricultural imports could not result in widespread famine.