Because having a copy of a gene is only one possibility of compensation, in processes like innate immune response it is common to have more than one pathway of activation, and to a certain point the presence of one protein can compensate the lack of another even if they are not structurally similar. A blast search can't be used in that case to rule out the need of that specific gene in other people. I am simply saying that proving that some people can survive without some genes without really having studied them to a certain degree is not enough to say that is apparently unnecessary for survival, it may be so only on certain populations (the sample in the studio is understandably biased).
That is of course even without going into genes that could not be necessary all the time but increase the possibility of survival to specific events, for all we know the sample of individuals could represent only the 10% that survive a gene deletion because of a specific diet, certain amount of exercise, lack of exposure to a type of pathogen not common on Iceland, etc.