Some minor notes to your article. I share your non-professional suspicion that rivers that won't be closed, maybe bays too, and - well - parts of the built-up Mississippi Delta would have to be considered, so the effective length wouldn't be so much shorter than for the Netherlands that is criss-crossed by the Rhine and canals and has lots of sub-sea-level land in need of multiple lines of protection.

I didn't found your $75 billion Iraq War reference at wikipedia. However, Costofwar.com, which is based on actual budget requests, is now at $192.5 billion.

_{*Lunatic*, n.
One whose delusions are out of fashion.}

Also, there's nothing costly about polderising sea. It's free. All you need is some wooden poles and time. Quite a lot of time, actually, but it works just fine.

But here I lost it:

If the current rate of erosion of these goes ahead, the sea will lap the levees of New Orleans this centrury and by the end the level of the land will be at least 1 metre further below sea level due to the "sponge" effect of extracting water while no river silt is deposited because of the existing levees.

That's one hell of a sentence and as I read it, it sounds wrong. I break it up in 2 parts.
1) Loss of wetlands will result in overflowing the levees
2)The "sponge" effect leading to lowering of the land is caused by extracting water and absence of depositing river silt

If that's correct with you want to convey, then I don't understand what you mean with the first point. Will there be overflow during a storm surge only, or in general?

Your second point is effectively solved by keeping a high groundwater level. The Netherlands have had a change in policy in this regard: pumping water out constantly increases the sponge effect, so they don't do it any more. The groundwater level is kept rather high to reduce land lowering. The risk comes during heavy rainfall, then the pumps need to get rid of the extra water fast. The absence of depositing river silt is only marginally important; it's groundwater that's the key.

My understanding of the hydrology of the river system is that it used to be a broad, very slow flowing river. When the river inundated the land that is now the lowest part of NO it allowed, as you say, the silt to precipitate out. That process was common to many major rivers and the Nile valley's productivity due to such flooding and depositing of nutrients built to country.

The main river at NO like most of its lower valley, is highly contrained by a system of levees and this does indeed increase the flow so main river silt is deposited way out to sea. Too far out to add to banks and swamps to build them up. NO also has canals and these alter the river's flow so that silt deposition can occur. With the river contrained in artificial banks, this reduces to total volume of water the system is able to hold. With nowhere to go but up storm surges make the tidal river overflow unless the levees are built higher.

If you compare maps of east London and NO  you will see both have large U bends where the rivers have meandered in the past. In London's case this formed the area of the main docks and, further back. wet reed beds. In those days, when the Thames had high water levels, it merely inundated the Isle of Dogs and took the shortest route. In NO that route is straight through the city.

Dutchman is right, the ultimate solution is not more and more constraints but allowing the river to manage itself naturally in areas where the costs of having floods are minimal. Just such a solution is proposed to protect the new developments in the Thames valley below the Barrier. As well as river walls, areas will be allowed to flood to absorb the excess water. Similar plans are already in place in East Anglia where the sea walls have been deliberately breached to allow the fields to revert to salt water marshes.