The authors talk about pull systems where production is driven by customer demand trickling up the supply chain. This is fundamental to modern low-cost manufacturing; for example it is part of Michael Dell's special sauce.
In software development you can do this too. Focus on delivering what the customer wants as soon as possible. If you can find out what the customer wants and give it to them quickly, you will have a happy customer!
In agile development, iterations are short and focused on delivering value to the customer. Each iteration should deliver the most important remaining feature of the project.
The authors give a quick summary of queuing theory. What I take away from it is that you get best throughput when you process in small batches and don't run your plant at capacity. The apparent optimization of running your equipment hard actually results in lower overall value for the process.
Again this points to short iterations. It also makes a good case for not driving your developers too hard :-) "Just as a highway cannot provide acceptable service without some slack in capacity; so you probably are not providing your customers with the highest level of service if you have no slack in your organization." (p. 81)
This topic comes up again in Chapter 7 in the discussion of optimization.
