[rationalbeats]

Well March is the Spring Equinox, and September is the Fall Equinox and that would be my educated guess as to why those two months are special.

[SoftwareMaven]

Serious question: Why would day and night on earth being equal length impact orbital mechanics?

Edit: I realized this might sound sarcastic. It isn't meant to be.

[ars]

It's not the day/night length.

It's the angle of the earth relative to the orbit of the moon. It's a geosynchronous orbit, so it must match the equator, including the angle.

Presumably in other months the angle of the moon's orbit (relative to the equator) is wrong for using this maneuver.

[btilly]

There are 3 bodies here, the Earth, the Moon and the Sun.

During the equinox the equator of the Earth lines up with the Sun-Earth axis. I can see that being significant in all sorts of ways. At other times of year, in a complicated maneuver, the Sun is going to pull the satellite off of that plane of orbit.

[eternauta3k]

From the patent referred to in the article:

"The present invention is directly applicable to a quasi-GTO having any given inclination but is especially advantageous to a satellite launched in either March or September into a highly inclined quasi-GTO because the lunar flyby is capable of removing all of the inclination"

Not very helpful. From another patent ( http://www.patents.com/us-6149103.html ):

"If the satellite 10 had been launched in March or September (around the time of an equinox), the lunar flyby could be timed to occur when the node of the moon is close to the node of the quasi-geosynchronous transfer orbit, so that the earth return orbit has an inclination near zero and the inclination of the orbit of the satellite about the earth could be completely removed by the lunar flyby. Launching March or September is advantageous for satellites or spacecraft because of the sun being normal to the required attitude of the satellite."

[Sharlin]

It's orbital mechanics impacting the length of day, not the other way around :)