The issue I have with this type of question is that it assumes the transaction costs to switch between tasks is 0 or minimal. If so, then yes, huge efficiencie

The issue I have with this type of question is that it assumes the transaction costs to switch between tasks is 0 or minimal. If so, then yes, huge efficiencies can be gained by figuring out where people are underutilized when and making use of that capacity. If the tasks are all similar and easy to switch contexts (i.e. the call centre in the article is probably a good example?), then OK, makes sense. But for a lot of things, switching contexts is expensive. Productivity drops, and any cost savings can be easily nullified. One may argue that the capacity wasn't going to be used anyway, so it's a waste if not used. But I find the calculations used to estimate capacity assume that transaction costs for switching contexts is 0 instead of some large number. Thus, they actually overestimate capacity, thereby causing not only lowering productivity because they jam too many contexts into the space that still need to be handled on time, but also lower job dissatisfaction.