I suspect that most of us assume that there is a linear relationship between the time we spend working and the amount we get done. I propose that this actually breaks down and at some point we damage our productivity by working beyond our optimum. Here is my oversimplified model of productivity.
Imagine you have a task where you need to move the maximum number of bricks from one pile to another. While there is no chance that you can move all the bricks, you would like to move as many as possible. How many hours a day would you work? If you can move a certain number of bricks in an hour, you would expect to be able to move twice as many in two hours, four times as many in four hours, eight times as many in a good eight hour day of work. It is tempting, since you'd like to move the maximum number of bricks by the end of the year to extend your day to maybe 10 or 12 hours. At what point will working more hours reduce the total number of bricks that you are able to carry. Without sufficient rest, you can imaging steadliy growing weaker throughout the year and being unable to move as many bricks in an hour. Confronted with this slowing pace, you may work longer hours, further eroding your productivity and reducing the number of bricks you move each day. There is an optimum number of hours to work that allows you to move the maximum number of bricks. Beyond that optimum number we risk fatigue and injury that a night's rest won't allow us to recover from.
I suspect that knowledge work follows a similar limitation in the amount of productivity one can expect reaching a maximum and then decreasing if you work too much. Unfortunately, I suspect that many of us do not see that there is an optimum point and working beyond that point is counterproductive. The idea that the most productive thing that I could be doing is resting so that I don't exceed my personal optimum for hours or effort spent is counterintuitive.
I'm not certain where this maximum is, but I'm certain that I have exceeded it and damaged my overall productivity by trying to increase my productivity through more effort. Admitting to this limit seems like a subversive act, but it is my commitment to productivity that compels me to put my work aside on a regular basis. I would love to do more in my year than I currently do, but that means somehow creating a new curve that goes higher. I can't do that by working beyond the peak of my existing curve. In my first year teaching, I found myself working in inefficient patterns because by working too much, I had eroded my creativity and was in a rut. It was only after periods of rest that I could find a way to increase the slope of the curve.