Although I understand your sentiment of frustration, I'm not sure if that is just a goof stemming from mistranslation or unfamiliarity to the subject. The snippet is taken from a physical geography site; not a branch of study particularly rich in mathematics. In the Netherlands, the field of water management is dominated by civil engineers and mathematicians.

I do know that the Delta Committee and others have adopted a probabilistic approach, and that probabilistic methodology now provides the baseline for evaluating dyke safety - which was something I was doing earlier this year. It also is estimated that some 25% of all primary dykes currently still fail the safety test - which in the next round in 2011 will become even more stringent. This does not create major problems that have not been seen before (see recent Maas flooding) - and as DoDo writes upthread upgrades are in the work at all levels.

After the Delta Committee a National Water Plan was released in 2009 by the Balkenende government, which adopts major recommendations of the Committee - including plans for providing major rivers with space for large variations in river discharge, whereby the highest discharge scenario is taken as the baseline, and then a 5 - 10% extra is added.

After operating at the consultancy side in this particular branch, I have enough anecdotes that point at some flaws in the present regulatory framework of the Dutch waterboards, however, I have no doubts that the conservative and stringent safety conditions that need to be adhered to continue to make the Dutch safety system against flooding simply the safest in the world.