Sort of and sort of not.

The primary difference between the DSGE type view and the conclusion being discussed from Leijonhufvud is linearity vs non-linearity – and thereby single eqm vs multiple eqm.

Now Leijonhufvud does also discuss the fact that we may not have stable steady states, and that would lead to the endogenous instability you are discussing for sure.

But even when there are locally stable steady states that allow the economy to “rest” in the long run, there is no certainty that this steady state is unique and as a result even in the face of a “temporary” shock we may find ourselves moving towards a different steady state. Typical DSGE modeling would not lead to this – a temporary shock would not change the long-run eqm.

Personally I think that the difference between the “endogenously generated instability” and exogenous shocks is one of semantics. It is an important difference to be sure, but in one case shocks are being generated endogenously and in the other shocks are being generated exogenously. Treating the shocks as exogenous (instead of a potentially more realistic endogenous case) allows us to get an idea of how the shock functions and what path it moves through – which gives us a clearer way of understanding change.