To get a full understanding of the different test situations you must know what the judges looks after, and also understanding dogs mental behaviour. There are more detailed descriptions, but I haven´t found any in english.
The "lure-coursing" test the dogs lust for hunting, to run after a prey and engage it, shake it and so, it helps if a dog has this lust and have intresst in objects during training and learning. This same thing is also tested when the dogs should find and engage in play with the "strange-dressed human", here they see if the dog has the same intresst in bigger "prey", and if he will engage in play and tug-of war when he is away from his handler.
The purpose with the pop up dummy is not that it SHOULD look like a human. They will see how a dog reacts in a surprise situation, is he scared, overly aggressive, how long does it take before he investigate the dummy, does he need help, will he pass it a second time without remaining fear? There is not a single trait they look for here, it´s many things that rules how a dog reacts in a certain situation.
The purpose with the test is that all dogs should be tested in the same way, so it´s possible to compare inborn reactions between induviduals or in a litter. You can´t for example select between two bitches to your studdog, if they have been exposed to different testsituations with no strict rules.
When the police or military test their dogs, they do it to select candidates for their needs. The police use a similar test with no strict rules, but also in most cases with a stake-out test, to see the dogs fightdrive, and also an enviromental test on slick floors and so on.
The military has a test to select military-search dogs, they do some of the situations in the test, but there are also added differnt search-excersizes.
So it´s important to differ between how a selectiontest for service is done, and how this test is done to benefit breeders, so they could in an easy way evaluate their breedings, and use the test to change both positive and negative traits in a breed. The problems with adding a stake-out test to see a dogs fightdrive in this test is that only a few persons have the experience to do this, and they wouldn´t have time to test dogs all their free time. Second problem with this is that it´s hard to test a dogs fighdrive in the same way on all dogs, you have to adapt your behaviour to that particular dog, and you also have to recon with a dogs previous training, so it´s hard to see genetic differnces between two dogs in such a test.
I don´t think that there are any major faults with the test, because it have been used in a long time to test reactions that is needed in a service and workingdog with good result. Similar test has also been adapted in denmark, norway and finland, and other countries to have shown their intresst. It may not be 100% perfect, but it´s the best tool we have today to measuer inborn qualities, and there is a constant discussion how this test could be refined even furhter by the k9 gurus. All breeders of mals and GSD for service and sport here in sweden also agre that this is a good test, and how a dog reacts in the test says alot of the quality of the dog in practical work, dogs that shows good nerves and courage in the test doesn´t usually have problems with slick floors and other enviromental things a servicedog must be able to work in, and nerves and lack of courage seems to be the biggest problems in servicedogs. The breeders who don´t like the test are mostly showbreeders, who don´t want to show the public how their dogs do in the test.