I've come to believe that big software projects are like a marathon. Burning up your maximum energy every step of the way simply isn't going to yield better results. You need to create a pace and keep to it. Sometimes you speed up, sometimes you slow down, but saying 60 hours/week from start to finish is gonna be counterproductive.
I remember reading some studies on this, I think it was in the world of game development. 60 hrs/week delivers more than 40 hrs/week for the first few weeks, but then it reduces and after about 6 weeks, you're delivering less working 60 hrs/week than you were at 40.
Of course, there are lots of related issues, like are your employees getting enough sleep? Very varied work might tax employees differently. Is 8hr days, 5 on, 2 off necessarily the optimal pattern? Can people work more efficiently for longer if they're working on something they truly believe in?
