"One thing puzzles me and maybe you guys can 'splain it to me. The author says there will be a tendency for a growing moon to move closer to its primary as it sweeps through the existing gas cloud. He goes to some length to explain how all moons don't end up accreting with their primary. Now my problem with this is that I thought that a moon's (or a planet's) orbit resulted from the net orbits (vectors) of all of the things that went into building it. If you are talking about a gas cloud that is busy building a gas giant, and there are leftovers to build moons with, then the moons are being build of gasses that are already in orbit (not at rest - whatever that would mean). Wouldn't moving through this gas (capturing some) that was already moving in approximately the same orbit as the moon have only a minor effect on the moon's orbit?"

I think it depends, if the orbital energy of all the matter is mostly conserved, then you are mostly right- if you limit it the matter to a disc shape orbit with mass going in same general direction.
So, to keep it simple, we will limit to the matter to being confined to a disc shape: if plantary objects slingshot objects, to higher orbits then a plantary mass loses orbital energy. Now, a planetary mass can gain orbital energy if robs orbital energy by making that object go to lower orbit- this object might later be tossed out of the system by the plantary object [ending up giving net loss to plantary body], tossed out by another plantary body, stay in the lower orbit, or end up colliding with the star, whereas a object toss out of system may stay out, giving a net loss. With collision with plantary object the orbital energy is mostly conserved but you do get small losses in the form the impact energy.
Let's look at extremes, a star without matter being added, has net loss of mass by converting matter into energy and radiating it [it's also creating a solar wind] though a white dwarf, which is dense hot matter is merely shedding energy and it would also lose mass. Now if an object is losing mass by radiating in every direction [negating vectors] and is in an orbit [and all objects are generally in some orbit] that orbital distance should increase, conversely increasing mass should decrease orbital distance. What I'm saying is that if you add mass to object (and vectors of all this mass is "balanced") that will decrease the orbital distance. This increase/decrease applies to the solar as well as the plantary masses.
Note the above is just my opinion- it could be wrong [probably is:].