So, now we have metrics of what "slighted at work" does to work.

Just like the terrible outcomes of fraud elimination indicate non-zero fraud is preferred, I can see a similar nom-zero dissatisfaction at work the same.

Except now companies can calibrate how much dissatisfaction due to terrible policies can be done in response to lower work quality, and minmax that.

Reading between the lines, the paper is how to manage dissatisfaction in relation to work and cost. In other words, "How bad can we make it for the minimum work product needed?"