A reassessment of the "open model" of social account proposed in the mid 1980s: are some insights still useful for the current social accounting debate?