What he describes after the curriculum committee anecdote does not look at all like the planning fallacy to me. In the planning fallacy the required scope/effort/schedule is underestimated, whereas in the case of the recovery from a recession it is not even clear that the work done by the government speeds it up, slows it down, or has no effect whatsoever:

The reality, of course, is that the economy is not a patient. It is a zillion, zillion interactions. Government is not a doctor. Most of the time, it is a clashing collective enterprise that is occasionally able to produce marginal change, for good and for ill.

An "economic recovery project" has no clearly defined goals beyond the obvious implicit one of having the participants reelected, and in that sense it either succeeds or fails in the next election cycle, without ever going over time.

As is the norm in politics (aka the mind-killer), all layers of government are driven by an eclectic mix of many cognitive biases on the part of the government itself, the voting public and the lobby groups. It would be a flattery to compare what they end up doing with even a poorly run industrial or academic project.