Food scarcity is mainly a problem of poverty and inequality.
. . . Enough poor harvests at the world’s major food producing regions will increase prices. The rich world will be fine but the world’s poor and middle income countries will suffer economic hardship if not malnutrition. This would have political consequences for everybody.
Quite so. Available data on decreasing cost and decreasing volatility of price over history might lull one into a false sense of security about global food adequacy. Until the 20th century, periodic famines were the norm, not the exception.
Large-scale farming has brought efficiencies of scale and lower prices. Transportation advances have evened out prices across the globe. These are the advances that have lowered both price and volatility of price. Neither of these, however, improve total global food production adequacy in any given crop year.
Food is an extremely "inelastic" commodity, in econ-speak. People don't eat much more food just because it's cheap, nor eat much less when it's expensive. Supply of food responds only slowly to changes in price. Price is thus a poor measure of adequacy of supply, if one is assuming low price means plenty of reserve supply.
If the total global production drops precipitously in any given year (drought, floods, fire, UG99, other pathogens, civil strife, whatever) prices can skyrocket. Those high prices will result in more acreage being planted, but there's a growing season of delay between changes in price and production. There's plenty of time there for lots of poor people to starve to death.
I think it's plainly true that affluent societies need not fret much over food availability. The poorer half of the world will starve before affluent societies feel any real pain.
Wheat, rice, corn, and soy are the individual measures to look at in terms of risk of mass global starvation. "Food" includes lots of other things that are fairly irrelevant in terms of global risk to life, even as (dollar-wise) they contribute quite a bit to totals. Compared to other goods, these staples still show quite a lot of variability of price, indicating that there's not actually a lot of reserve of supply over demand.
So yes, yields in Africa are quite a bit below optimum. But the individual farmers are doing their best with their resources at hand. They'd be crazy to let production lag behind their possible yields, if they had any choice about it. Raising those yields is a multi-year, pricey prospect. This isn't a resource that can be tapped in the face of a famine of a couple of years.
I think, overall, it's plausible that the world can feed even a somewhat larger population for several decades. I think, however, that it's about as plausible that catastrophic global starvation for the poor could happen in any given year.
I think sane governments should prepare for the "seven years of plenty, seven years of famine" right now.