This is definitely tricky to manage for an Earth-Moon-Sun like system.
The daily motion of the Sun is generated by the rotation of the Earth. The seasons are caused by the orbit of the Earth around the Sun. The axis of rotation of the Earth is inclined at about 23.5 degrees to the Ecliptic, which is the plane of the Earth's orbit. The orbital plane of the Moon about the Earth is inclined at about 5 degrees to the ecliptic, which causes the Moon to go off the Sun's path in the sky, and we don't have eclipses every time they pass.
If planets condense from a debris cloud, this tends to give orbiting bodies with a common plane. Most of the orbits of the planets are close to the ecliptic plane, and to a lesser extend the axes of rotation are perpendicular to the ecliptic (noteworthy exception with Uranus). This suggests that if the Earth-Moon system could condense from some debris cloud at a much smaller scale than the Earth-s orbit, then it would be possible for the Earth-Moon system to evolve with a tilted plane.
There is a more subtle effect, called the Kozai-Lidov effect. You can have an Earth-Moon system with a tilted plane, but this can exchange orbital inclination for orbital eccentricity, and back again. Above an angle of about 40 degrees, this means the orbit eventually becomes so eccentric that the two bodies hit each other.
So, what about Uranus? That whole system is tilted over at 90 degrees. The answer is that the Kozai-Lidov effect gets a lot weaker with distance, so what would cause a binary planet to crash in a year at Earth's orbit might take many millions of years for an outer gas giant. Even if your moon is tiny, like Phobos about Mars, and is probably a captured asteroid so it can have any orbital plane it likes, it is only still orbiting because its orbital plane is close to the ecliptic.
The Earth-Moon orbit plane is probably pretty close to the ecliptic. You moon might be below the ecliptic for half a lunar month, and above it for half a lunar month. Our Moon is probably hidden below the horizon at winter at high latitudes for a few days a lunar month. If you want it to stay hidden for seasons, you want your lunar month to last a year. You can do this by moving the moon about 6 times further from Earth. If you want it to look the same size, then it will be bigger, so the system is almost a double planet.
Can you have a lunar orbit with the same period as the solar orbit? Laplace resonance is for planets with a common orbital centre. I feel a lunar orbit would not 'lock in' to the solar orbit, but be destabilised by it. So the Lunar period would be different, but it could still be something close to a year.
There is one other place a moon might hide. It is pretty extreme, but Phobos does it so I mention it for completeness. If you are close to the Martian poles, you can't see Phobos because its orbit is in so close that it is hidden by the curvature of mars. But Phobos is almost on the Roche limit, so it is almost breaking up from total forces. Its orbital period is about 7.5 hours.
There is one possible solution if the moon is a lot smaller and lower. If the moon was in geosynchronous orbit, it would be permanently hid from one side of the planet; if it was in near geosynchronous orbit, then it would be hidden and then reappear. I suspect such an orbit would be unstable on astronomical timescales.