As our knowledge base increases, so does the need for the scholarship of integration and synthesis. Reviews by method are key for novice researchers and those returning to active scholarship after a period of time off for whatever reason. Reviews by subject matter are key for textbook writers (at least those that are evidenced based rather than standards based), regulators, standard setters and practitioners.
But standards for reviews must improve and editors need to enforce those standards. I recently saw a paper rejected by one journal and published in another, with all the problems that lead to its rejection in the first journal still present in the second. The biggest problem, you could not replicate the review based on the authors’ description. Maybe they did a thorough search of the literature maybe not. We simply cannot tell. They did not disclose what journals they included (or more importantly excluded albeit you could guess based on the journal cited), how did they search to find the articles (i.e. what were the inclusion/exclusion criteria), what was the population of articles they found and how did they narrow it down to what was included in the review. And that was simply the start of the problems.
Now if you are interested in a “biased” literature review based on the “great man” theory of selection of literature – these are the reviews for you. However, in most other academic disciplines they would get laughed out of the community and any journal that published such an “unscientific” review would be regarded with disdain.
The bottom line is there are standards for reviews just as there are standards for almost everything else. Editors need to obtain competence in this area just as in any others. Being a journal editor should not be solely a “learning on the job,” new editors need to be mentored and given feedback. They have to know it is okay to ask for help from their more experienced brethren. To do any less is to devalue the research that is being reviewed!!!!