No, not inhabited satellites.
Mainworlds.
As defined on Page 37
LBB:6 Scouts and in other locations in all other versions:
The main world is the world in the system which has the greatest population.
We've had inhabited satellites, planetoids, KBOs, dwarf planets, flotsam, and jetsam since
LBB:6. The question is how many mainworlds should be satellites of larger bodies. Answering that question is hard for two reasons.
First, none of
Traveller's various sysgens are scientifically accurate. This isn't because those sysgens are "bad" or the people who wrote them are "stupid". It's because the actual science is in such a state of flux. Every exoplanet discovery shuffles the deck supporting some theories, discounting others, and adding more data which needs to be taken into account. Sysgen is a mess because the real world science it should be based on is a mess. An exciting mess, a mess in which wondrous discoveries are seemingly made every day, but still a mess.
Second, the only ways a mainworld can be a satellite is either "by hand" or as a result of a series of very low probability rolls in sysgen. The "by hand" method produced the laughable distribution on the map sudnadja kindly shared. Writers and others over applied a trope they felt was "cool" and crammed dozens of
Sa mainworlds in a few sectors.
LBB:6 sysgen can produce
Sa mainworlds, just not at the rate seen on sudnadja's map. Even the Marches' two examples may be too many when using
LBB:6. Coming up with an
Sa mainworld in
LBB:6 is convoluted because it requires several things to happen.
You need to already have the mainworld UWP, you're using the continuation system. Next, the system you roll up needs to have enough non-mainworld components and enough empty orbits to "force" the placement of a gas giant in the habitable zone. When placing known components, gas giants are placed first, belts second, and an
existing mainworld last.
Wil could probably run the probabilities in his sleep but the chances of having a certain number of gas giants which can only fit into a certain number of orbits if the habitable zone is used can be calculated. My back of the envelope jottings suggest that the Marches' two systems is roughly the expected amount.