Questions about a habitable second moon

[Bandersnatch]

Let's say we start over? My restrictions are that M2 is comfortably habitable, has a 32-hour day, and is tidally locked to P. P (inhabitable) still has M1 and M2. (Bonus: M1 was habitable but now is not.)

Potential habitability of the satellite, to a good approximation, should be determined by the distance from the star. I.e., if the parent planet is in the habitable zone, so are its satellites.
Other reasons than solar irradiation can still affect habitability, though. So you can e.g. say M1 used to be habitable, but being small has lost too much of its atmosphere by now. While M2 being larger has retained its atmosphere and is habitable just fine.
This also suggests the parent planet should not be an ice giant, as the ices should melt/sublimate at the distance from the star where the moon can be habitable.
Energy from tidal stretching or otherwise received from the parent planet could alter this picture, but IMO it's better to ignore these for clarity.
Similarly, I'd handwave away any radiation concerns. You can always say it's remarkably low, or the moon happens to orbit away from the main belts, or itself has a strong magnetosphere.
Or, you could say that while M2 orbits away from the radiation belts, M1 has at some point drifted (due to tidal interactions) into one of the belts, which accelerated atmosphere loss and is the reason M1 is uninhabitable. Maybe it still has enough atmosphere to survive on, but it's now too strongly irradiated to support life? There are options here.

Here, use this spreadsheet.
(Google Docs; use the hyperlink and make a copy to edit - otherwise it's viewing only)

There are two sheets in there.
The first is for the star and its habitable zone. The data should be left alone here, unless you really need more/different masses. But it already spans the range of masses corresponding to G and F-type stars. If you really want to, the masses can be extended on either end by a bit, and the table should still make sense.
Pick a star you like, and an orbit within its HZ, and copy the mass and the desired orbit into the second sheet.

The second one is for the planet-moon system. It's a bit more messy (sorry), just remember that green fields want you to enter something, while blue fields tell you something. It calculates things like the sizes of the two planets, how large the parent planet would look on the sky, if the orbit is not too large for the moon to fly away, or what the period has to be given the radius (or vice versa). A few other things.
The data already in place is for a Sun-like star, Earth-like orbit of the parent planet, a Uranus-like parent planet, and an Earth-like moon tidally locked on a 32h orbit (which sets the length of a day in this case).

There's precious little room for another moon on an even lower orbit, so either raise the orbit of M2, or place M1 on a higher orbit than M2. (M1 is not included in the spreadsheet)

There's nothing there about orbital inclination or axial tilt (yet). If you decide to have axial tilt AND tidal lock, then the orbital inclination will have to be equal to the axial tilt (and vice versa). Without tidal lock the tilt and the inclination can be uncorrelated.

With the current setup, and without inclination, significant daily eclipses would be commonplace. These are not taken into account when deciding habitability, or anything else for that matter.

Try playing with it a bit and see if any of it makes sense. Some cells have comments telling you what's what, but let me know if you need clarification or want something added that you can't do yourself. Or if the link doesn't work for some reason.

Also, I can't guarantee everything's put in correctly. Comments welcome.

 

[LadiSilverfox]

That's going to influence my visualization of the planet, somewhat.

How much scientific accuracy is necessary for the storyline ?

Basically, I am trying to figure things out like how often they see M1 from M2, if possible, in what phases. When/How often M2 experiences eclipses or other such events, it will tie into how certain outer spiritual realms will interact with M2.

 

[snorkack]

Consider then the three systems of densely packed big satellites in Solar System.
Jupiter: 317,8 Earth masses
Inner large satellites:

1. Io 422 000 km +1,7627 d
2. Europa 671 000 km +3,5255 d (2:1 resonance to Io)
3. Ganymede 1 070 000 km +7,1556 d (2:1 resonance to Europa)
4. Callisto 1 883 000 km +16,69 d (out of resonance)

Saturn: 95,2 Earth masses
Inner large satellites:

1. Mimas 186 000 km +0,9424 d
2. Enceladus 238 000 km +1,3702 d (not resonant to Mimas)
3. Tethys 295 000 km +1,8878 d (2:1 resonance to Mimas, not to Enceladus)
4. Dione 378 000 km +2,7369 d (2:1 resonance to Enceladus, not to Tethys)
5. Rhea 527 000 km +4,5715 d (out of resonance)
6. Titan 1 222 000 km +15,9454 d (out of resonance)

Uranus: 14,5 Earth masses
Inner large satellites:

1. Miranda 129 000 km +1,4135 d
2. Ariel 191 000 km +2,5204 d
3. Umbriel 266 000 km +4,1442 d
4. Titania 436 000 km +8,7509 d
5. Oberon 584 000 km +13,4632 d

no resonance for Uranus.
Your quoted 32 hours for M2 is close to what Enceladus and Miranda have. And there is the major satellite Mimas inwards of Enceladus, as well as minor satellites inwards of all of the aforesaid.
What do you mean by "twice faster" for M1? Velocity? Or maybe period, as in being in a 2:1 resonance to M2? There are 4 pairs of 2:1 resonances among the 15 major satellites listed, so a likely if not inevitable arrangement.
What is "Uranus-sized" for your "silicate" planet P? Mass, or radius?

Also a question for the astronomers/celestial mechanics here...
Precisely what are these 4 2:1 resonances precise to?

This ties to the question of what phases will M1 be seen.
If M2 and M1 are out of resonance than M1 will be seen in all phases.
If M2 and M1 are close to resonance but not precisely then M1 will also be seen in all phases, but over a long time period.
But if the resonance is "precise" then precise to what?
Consider near side of Europa. Jupiter is visible at all times and all phases. In one orbit (3,5255 days), Sun circles the sky so half of the time it is night and some of the rest, eclipse. But at the same time, Io completes 2 orbits around Jupiter - thus catching up with Jupiter just once.
In a long term, do conjunctions of Jupiter and Io as seen from Europa happen always in the same constellations of fixed stars? Then they happen in all phases over 12 year orbit of Jupiter. Or is the resonance exact in synodic terms, so the conjunctions are always in the same phase? Or neither?
Same question about the other 3 pairs (Ganymede-Europa, Tethys-Mimas, Dione-Enceladus).