Flight schedules are an art as much as they are a science. I heard a guy from CO give a talk on that issue once, and he said the ideal published flight times are such that 50% of the flights arrive early and 50% arrive late. Too much on either side and you create yourself problems. Part of the problem is that there are some costs directly tied to published flight times, such as pilot pay. Pilots are paid based on what is in the schedule, or actual flight time, whatever is greater. So, if the flight arrives early, the pilots have to be paid "extra." Sitting around on a taxiway, waiting for a gate, is burning fuel unnecessarily. On the flip side, if the flight is late, pax may misconnect which is not a pleasant experience for anybody. The point is, early is bad and late is bad.

Old NFO is right about aircraft altitudes. He is speaking in generalizations, and the reality of the matter is that based on a given weight and atmospheric conditions, there is an optimum altitude and speed to fly for using the least amount of fuel burn. The service ceilings for an airliner are typically in the low 40's (40k feet) and flight paramters are typically not optimum at the service ceiling. In fact, as you approach the ceiling, it gets harder and harder to fly, increasing fuel burn among other things.

The other thing about performance calculations is that flying slower doesn't automatically mean less fuel burn. Two things come into play: Slower flying means more airborne time, which means more total fuel burn, even if the fuel/hr burn is lower. Also, by flying too slow, you can get "behind the power curve" which means it actually takes more power to fly really slow. My point is that most of these calculations aren't linear, which complicates the performance calculations and means you can't extrapolate any of the performance numbers without having the manufacturer's published charts.