All five of these theories are best taught sympathetically by being taught historically: as long traditions of thought that smart people have used to try to understand a changing and confused world. Thus Minskyism from its nineteenth century roots with Walter Bagehot or perhaps Adam Smith grappling with nineteenth-century financial crises, Keynesianism from its roots in Knut Wicksell's studies of disturbances to the flow-of-funds, monetarism from its roots in John Stuart Mill trying to understand the first industrial downturn in England in 1825, overinvestment theories from their roots in Karl Marx grappling with the crisis of 1848, high-real-wage from its roots in Nassau Senior's examinations of technological unemployment in the pre-1850 Midlands--all tussling with a set of problems first raised by Jean-Baptiste Say and Thomas Robert Malthus.
A couple years before the crisis, the GMU econ department asked if I wanted to teach a macro course. I outlined a "history of thought" course for macroeconomics. So I agree with DeLong that this a useful approach for teaching it.
However, I think that the fundamental question of macroeconomics ought to be posed as follows:
Why are there fluctuations in the amount of time people spend on market activities and non-market activities?
Note that this year's Nobel Prize was awarded for suggesting that the non-market activity is job search. This may be one way of describing the process of groping for new patterns. However, I do not think it is necessarily the theoretical approach that will prove most satisfying.
Instead of my one question, the paradigm of aggregate demand and aggregate supply creates two questions. (1) How does money or finance or investment or savings shift aggregate demand? (2) Why is aggregate supply not vertical? Several of the approaches that DeLong lists do not address the second question. Interestingly, Keynes himself cannot be pinned down to a precise answer to either question.
As you know, I would prefer to start without the aggregate demand and aggregate supply paradigm. I see economic activity as patterns of sustainable specialization and trade. My basic explanation of economic activity begins with Smith on specialization and Ricardo on comparative advantage.
This leaves the following puzzle: the gains from specialization and comparative advantage appear to be immutable; why would market activity ever decline?
My attempt at an answer is that while the gains from specialization and comparative advantage are immutable, the pattern of trade that achieves these gains is far from obvious. The most reliable pattern today is usually something that is very close to what worked yesterday. However, if a lot of yesterday's patterns are not sustainable today, then entrepreneurs must grope for new patterns. That can take time.
DeLong presumably would place the PSST approach in the tradition of theories of malinvestment. I think that is right, provided that we recognize that malinvestment includes malinvestment in human and organizational capital, and provided that I be allowed to appropriate Minsky-esque sources of malinvestment, so that I am not limited to malinvestment that is caused by monetary policy.