According to the discussion of Baddeley's approach, working memory is not just a passive storehouse. Instead, it is like a workbench where material is continually being handled, combined, and transformed. Why is the workbench metaphor more relevant for Baddeley's model than for the Atkinson-Shiffrin model?