I think that this has to deal with "grey" litterature. Some references are missed becaused either their keywords do not fit with the research strategy (lack of sensitivity of the search) or some journals that are not archived in the databases that were assessed. This is why other possibility to look for possible new study to include can be performed using either the method you mention or also asking key researcher on a particular area to look for this type of information.
however, does all these supplementary efforts make a difference ? I think it is question dependent as well as depending on the way traditional bibliographic database assessment is performed.
This is because within the literature that has been searched, there may be references to other valuable resources that happen to be outside the scope of the search strategy, timeframe, or databases used.