I think his point is that large ambitious projects tend to fail by their very nature.
The budgets are huge, the stakeholders are legion, and no-one can agree what they really want, because they all want entirely different things (most of which are orthogonal to the project at hand and to do with careers, ego, and corporate politics). It's very hard to make a project with a long timeline and a large budget come in on time, let alone one which involves three massive bureaucracies and management chains, even if the functionality to be delivered is simple.
The budgets are huge, the stakeholders are legion, and no-one can agree what they really want, because they all want entirely different things (most of which are orthogonal to the project at hand and to do with careers, ego, and corporate politics). It's very hard to make a project with a long timeline and a large budget come in on time, let alone one which involves three massive bureaucracies and management chains, even if the functionality to be delivered is simple.