The Source Material
Very recently (i.e. today, as of posting), I've realised there is another error in the 2300AD near star list. It is not the first, and several of these errors have become necessary for the universe (such as Queen Alice's Star). This is one of those.
Ellis orbits Gliese 623. This is now known to be a binary system with two red dwarves orbiting their mutual barycentre 1.9 AU apart. This means no planets can exist in the notional habitable zones. However, the second star of the pair wasn't reported until 1989, and the Gliese catalogue GDW used lists the star as an M3V, i.e. a red dwarf (by the standard tables, 0.36 solar masses and 0.016 solar luminance). However, look at the 1st edition NSL entry:
It is listed as a G0V star (1.04 solar masses, 1.21 solar luminosity by the GDW tables), not a red dwarf.
The Colonial Atlas entry is by William Connors, and says the star is a M3VI, with a mass of 0.013 solar masses and a luminosity of 0.0065. This is a problem. The listed mass is 13.6 Jupiter masses, i.e. a brown dwarf. It is not either the NSL yellow star, or the source material M3V. The luminosity is a nonsense, as the BD is just on the limit of deuterium fusion, and this would have been expended very early in the BD's life (about 10 million years after formation), leaving a large gas giant.
There is no possibility of any habitable planet in such a system. Ellis is listed as orbiting at 0.07 AU (~ 10 million km), which means it would get 1.33 times the insolation of Earth whilst the deuterium lasted (using the write-up luminosity), but then would go dark. It would entirely be infrared to dull red, with no green or blue light to sustain photosynthesis.
In a further wrinkle, the 2300AD Resource, which is also considered canon, places Ellis as 0.013 AU from the star, and the shelf at 0.28 AU, which makes the stellar mass 0.08 of Sol. This would be an M9V star, large enough to sustain hydrogen fusion, giving an insolation of 0.05 of Earth.
Then there's the 1988 boxed set, which says:
Yet another stellar type! Now Ellis' star is a G8V (i.e. a yellow dwarf slightly smaller than Sol, 0.93 solar masses and 0.63 solar luminescence by standard tables). Yet, and the same time the NSL entry in the 1988 edition was edited to match the Colonial Atlas writeup!
The Fix
What to do in this bizarre situation?
Firstly, let us observe that every single listing for Ellis' star type is wrong. Secondly, we cannot use the real system, because a planet is impossible. Thirdly, we want to sustain the existence of the planet as described in "Liberty."
The best answer to to accept the 1988 writeup, and have that override Connors' mess. The planet Ellis would orbit about 0.8 AU, giving similar insolation to Earth. Oyster would be about 1.6 AU, and Gibbet about 3 AU. The FTL shelf is about 2.36 AU out, and so visiting Ellis, which is too dangerous to orbit, means taking a ship to Boise (near Oyster), and then making a 0.8 AU journey. This would be perfect for a stutterwarp equipped lander, and would justify America building such things.
