As for the article itself, as it is from harvard business school, the author may indeed be concerned about margins. I'm more interested in the inherent dangers of a system that makes you extremely vulnerable in bad times, and how to balance that against the obvious advantages, in good times, of that system.
As to war, what is the probability of the countries actively tied up trade with each other going to war. Obviously the US and China are posturing all the time, but war?
Plus trade and countries are dynamic and flexible and changing all the time even without major external shocks. A sudden reduction in supply in location X, would shortly lead to a rise in supply from location Y.
I only mentioned drought, because, in an agriculture based economy, a drought is devastating, as you suddenly have a huge reduction in your exports. It doesn't even have to be what you're imagining. Just a wheat crop that's only 60% of last year, could be ugly.
Anyway, we know what the most likely disruption will be, and it's not war, it's currency collapse.
As the island only has 1 bridge connecting it to the mainland, and the ferry service is largely only a novelty service any more it would present a serious problem for the country for its food production.
I think this model works extremely well ... until it doesn't. Until some event keeps trade from occurring, whether that event is a currency collapse, a drought, or, more often, a war.
I don't know what the right answer is. We seem to be muddling along fine, however, with the system as it is, but there's always a specter of catastrophe hanging over our heads.