It comments that air launches reduce the costs to orbit not (as ColinWright rightly points out) through energy savings but because it 1) doesn't have as many weather-related launch delays, 2) it can insert into almost arbitrary orbits, 3) there's no need for a blast-proof pad and related ground equipment, 4) launches over an ocean save on insurance costs, and reduce collateral damage should an explosion occur, and 5) the higher altitude means the first stage engine bell doesn't need to optimized for higher pressures, and the lack of high cross-winds means there's no need for gimbals; lighter fins suffice.
On the other hand, it also says that Pegasus is "one of the most expensive "launch-to-orbit" vehicles" (I assume per-kilogram to orbit), but that the flexibility in choosing the orbit makes up for it for small payloads which otherwise must piggyback.
Not to mention that this dramatically reduces MaxQ (the maximum dynamic pressure on the airframe), which reduces design constraints on the spacecraft/upper stages of the rocket.
I hadn't realized that there already is an air-launch-to-orbit system called "Pegasus" (http://en.wikipedia.org/wiki/Pegasus_(rocket) ).
It comments that air launches reduce the costs to orbit not (as ColinWright rightly points out) through energy savings but because it 1) doesn't have as many weather-related launch delays, 2) it can insert into almost arbitrary orbits, 3) there's no need for a blast-proof pad and related ground equipment, 4) launches over an ocean save on insurance costs, and reduce collateral damage should an explosion occur, and 5) the higher altitude means the first stage engine bell doesn't need to optimized for higher pressures, and the lack of high cross-winds means there's no need for gimbals; lighter fins suffice.
On the other hand, it also says that Pegasus is "one of the most expensive "launch-to-orbit" vehicles" (I assume per-kilogram to orbit), but that the flexibility in choosing the orbit makes up for it for small payloads which otherwise must piggyback.