Initially Congress thought they had asked for a protection level that the Corps could provide for under $100 million in a decade or so. By the time all the studies were in they were looking at a billion dollar project stretching over forty years. They hemmed and hawed, chopped a little here and there just like any manager would and approved. You can bet they extracted a little hide in the process. Managers were called in for meetings and hearings. Promises were extracted. Pledges were made. By 2005 the project was way over due, 60-90% complete in areas according to the GAO and not expected to be complete until 2015. The Administration had been redirecting funding to Iraq and the managers were frustrated.

In the meantime flaws in the design began to emerge and time marched on.

• Meterologists were doing a better job of storm prediction and began to see some cycles they hadn't seen before. In the most intense cycle the chance of a "Project Hurricane" was much greater than 1% and a more intense storm could be reasonably expected.
  
• The NGVD published in 1929 was updated in 1988. Louisiana was sinking so the benchmarks were all wrong. In 2005 the 1988 benchmarks themselves were 17 years old.
  
• Computer based hydrological modelling technology advanced each year and the marsh around the city changed. The ecological impact of the MRGO and the oil field canals was not factored into the original models.
  
• The designs used by the Corps had safety margins that can be precisely determined using Civil Engineering methods. For some reason the margins the Corps used were in the 1.3-1.5 range. Nowdays we'd consider those margins suitable for a rural area. A city deserves more like a 2.0 or better design safety margin but that would be much more expensive.
  
• Safety margins depend on the strength of the soils around the levees. The Corps tested the soil in some locations but consistently overlooked testing the strength at the critical "toe" of the levee. Well we sure know the toe is critical now and some engineers may have thought it was critical in the past but were they ignored?
  
• Tests done in 1986 unearthed a critical flaw in the I-wall design favored by the Corps in the New Orleans area. Although the results of that test were widely circulated, the construction of the New Orleans floodwalls proceeded in 1992 with the faulty design.
• Levees built out in the swamps began subsiding almost before their construction was complete. No funds were available to maintain their specified heights so they just sank.