Ever since I started GMing, it's been my style to think of my campaigns like TV series. I try the best I can to fit an 'episode' (one full scenario, in normal RPG terms) into a single session and aim to wrap up a 'season' (a single story arc) within a set number of episodes.
This is in part because the first game I actually ran was the Angel RPG and the GM advice for that game framed campaigns that way, so it's what I'm used to. The other part is that I have mostly run campaigns for university-based societies, first with the Stirling University RP society and, more recently, with Glasgow University Gaming Society. The 'TV-style seasons' fit reasonably well with the semester calendar, I run the first half of the season in the first semester and wrap up with a mid-season finale just before the Christmas break, then continue the campaign when the society starts up again for the second semester, aiming to run the season finale just before exam fever kicks in. Then, over summer break, I might run a shorter 'season' of a different campaign, just like TV networks sometimes do short-run shows between seasons of their main shows (like NBC have done with Agent Carter during the mid-season break of Agents of SHIELD). There are some difficulties I've encountered with running campaigns this way though.
First of all, it's not always possible to fit a full scenario of RP into a single session, so sometimes I need to carry the scenario forward into the next session. That's not a big deal if it only happens a couple of times in the campaign, they'd basically be the two-parters that you get maybe once or twice in a TV show's season. But due to the unpredictability of PC reactions it's possible that the scenario will take longer still. Plus, as was the case with my Firefly campaign recently, you might wrap up a three-part scenario, only for the next scenario to wind up carrying over to a second session as well.
Second, trying to fit scenarios neatly into single sessions of play means you have to tightly pace those sessions and find a balance between plot-centric scenes and character-based scenes. This is a balance I haven't quite managed to perfect with my own sessions, and I generally find that when I allow for character based scenes it adds to the running time of the scenario in such a way that it usually has to be carried over into a second (or even third session).
Finally, running a full-length season arc (16-22 sessions) is pretty daunting and there's a real danger of losing steam halfway through the arc. Having the Christmas break in between the two halfs of your season might help with this by giving you breathing room at the halfway mark to revisit your notes and figure out the direction you're going with the second half. Or it might drain away whatever momentum you had built up already, meaning you have to build it up from scratch when the campaign resumes after the break.
So, what are some solutions to these issues?
With the first and second, the issue might simply be one of session length. Due to the time we have the gaming space for and the varying schedules of my players, we usually end up starting at 7:00pm and ending at 10:30pm. That gives us three and a half hours to fit an episode's worth of RP into. TV shows generally follow a four-act structure; if we take our cue from Matt Wilson's Primetime Adventures, an act might equate to half an hour or more in game time. However, the way Primetime Adventures is structured and the way the mechanics of that system work are specifically designed to streamline play and make it fit the TV show paradigm more closely, so let's extend the length of an act in game terms to 45 minutes, maybe an hour. So if we started at 6:30pm, we'd potentially be able to fit the scenario more neatly into a session. That is, assuming I also watched the pacing of the session to make sure it was progressing at a rate of 'an act per hour'.
There are another couple of tools I recently discovered which might help with the second issue. First off, taking a cue from Monsterhearts and asking players in between 'plot scenes' if there is a scene they would like to play out with one of the other characters. I did it once with mixed results, but then my players weren't expecting to be given the opportunity to frame their own scenes, so perhaps if I continue to offer the option they'll grow more accustomed to it and it'll eventually spark more character-focused scenes.
The other tool, which can be used in conjunction with the above is 'The Talk', a technique recommended by Quinn Murphy as part of the IndiePlus 2014 Anthology. If you haven't read it, you should give it a look, because it's a really cool technique for framing scenes where the main focus is conversation between characters. I've been meaning to distribute it to my players in the Firefly campaign and encourage them to try it out in downtime scenes, but I haven't got around to it yet. Sadly, it's probably a bit late in the game to be bringing these ideas in, since we only have about three sessions left in the campaign, but it's something I'll keep in mind for future games.
As for the last issue, the solution might lie in this article by the Chatty DM, which is simply to run shorter 'seasons'. In this way, the first semester of play would cover one short story arc of 6-8 episodes/sessions, and if the players wanted to continue after that, then the campaign could continue into a second 'season' of play after the Christmas break. So the GM gets the benefit of taking a breather over the Christmas break, and the need to regain momentum isn't as big an issue since they're starting a new story arc anyway.
As I've said, it's a bit late in the game to worry too much about applying these fixes to my GMing style to Firefly, but they're some things to keep in mind and I hope it's been useful to you in some way as well. If not, apologies for rambling on.
See you later!