Yes, they're missing the point Greatnan. The risk factors were compiled in many different countries over two decades, then a meta-analysis of the re-occurring issues was used in North America, UK and Australia to develop testing and profiling by computer on the VISOR (Violent and Sexual Offending Register) police system. These risk factors like 'never been in a relationship for more than two years' or 'lost job because of disciplinary matters' have proved to be relevant in a significant number of cases (although Peter Sutcliffe would have got through the test relatively unscathed!) A clinical override has to be used throughout, to be sure the tail isn't wagging the dog and things are overlooked. This system was developed after Soham, and it's extremely sophisticated now. For example, the police will have fed in the given details about the time of day Alice was seen getting in a van, information about the vehicle, and anything else that can be profiled, and the suspect and several others, and their known vehicles, would have been thrown up for investigation in seconds. Each time more info comes in, like the possible places to look suggested by the public, they are fed into VISOR and more suspects will be identified. Of course, you only get out what is already in the data. Some offenders are not yet known to the authorities.
For some papers to condense all this into damning a sector of society because they are deprived, unemployed or not in a relationship is potty!