The present economic trends are a case in point. Notice that the economists who attempt to make job predictions are consistently wrong. This at least suggests a flaw in the theories they apply.
Then relativity came along, and showed that Newtonian physics failed in some extreme circumstances (particularly high velocity). The theory was refined by Einstein for those extreme conditions. This did not affect the theory for routine cases.
As you so aptly point out, Newtonian physics breaks down in extreme conditions. The economies are far larger today than was the case in 1823. The speed and efficiency of the transfer of capital and information has also changed. My premise is that as the economies represented by Ricardo's underlying assumptions have changed from those he was familiar with, the theories he promulgated may have become inaccurate.
Who gives a flip about predictions? I care about what actually happens. And by historical standards of the last fifty years, unemployment is low and job creation numbers are good.
Economists do not use Ricardo's theories to predict employment numbers. They have econometric models predicated on a variety of theories, many of them quite recent. You can critique those if you like, but you can't pin job prediction problems on Ricardo.
My premise is that as the economies represented by Ricardo's underlying assumptions have changed from those he was familiar with, the theories he promulgated may have become inaccurate.
Fine, that's your premise. Where's your data?
So you think Ricardo may have become inaccurate because of the size of the economy and "speed and efficiency of the transfer of capital and information"? Hey, how about Adam Smith and the invisible hand? It is subject to the same changed conditions. How about representative democracy as a form of government? Same story. You can question anything as a result of changing conditions, but until you show some data to back that up, you're just blowing smoke.
From my own observations of economics over time, I'd say if anything that the complexity makes Ricardo's theories (and Smith's) even more applicable. In particular, as Hayek and von Mises added to the theory in the 20th century, one of the reasons laissez faire economics works is because no one person or small group can possess enough information to intelligently "manage" the economy. (The collapse of the Soviet system is the most obvious manifestation of this problem.)
So more complexity means it's even more important to keep away from such nonsense as "managed trade". It is literally impossible for anyone to gather and process enough information to make intelligent decisions about where to restrict trade in a way that the entire system benefits. Let me say that again. It's impossible to do what you anti-free-trade guys want to do, which is to somehow manage the process to protect certain jobs and industries, without having side effects that are worse than the benefits.