It seems Jade Reagent is breaking a few patterns.
The big dungeon with the McGuffin at the end is usually in book 5 (see: RotRL, CotCT, LoF, KM, SS). But in JR, book 1 follows this model. That's never been done before.
(Usually book 1s are pretty high on roleplay: RotRL, SD, LoF, CoT, KM, SS, CC)
The 2nd book is usually the "high concept" book that people have strongly divergent opinions about, either loving or hating it (see: RotRL, SD, CoT, KM, CC). But in JR, the 2nd book is just kindof holds it steady without breaking any new ground.
The 3rd book is usually the odd one out: either breaking with the AP's theme/plot or provoking GMs to ask what they can swap it out with as the poor reviews mount up (see: RotRL, CotCT, LoF, KM, SS, CC). But in JR, book 3 is the most on-theme and possibly most lauded chapter in the series.
Let's see what holds for book 4 (the sidequest), book 5 (the dungeon), and book 6 (the meatgrinder), and if any more patterns will be broken!
If you have to over-generalize and paint in broad strokes to make a pattern out of 9 data points, is it really meaningful if something doesn't fit into it? I think this is more like seeing shapes in clouds.
I mean, I would argue that second adventures are not necessarily "high concept," but often "skill based, possibly investigations." Jade Regent would fall right into that definition. So do some first adventures, or later adventures (See: Carrion Crown 1-5). They are a bit frontloaded because it is easier to pull such things off before magic gets too crazy. Not impossible otherwise, but easier and thus more likely.
Third and fourth books can be a bit sidequesty or off-theme, but again, so can all others. Such things tend to be loaded towards the middle for obvious plot structure reasons. Early adventures are needed for set up and late adventures are needed for pay-off, leaving middle adventures for sidequests or other odd things. Not to say it is really a "pattern," just "a bit more likely."
"Big Dungeon with MacGuffin" is broad enough to describe about half of all adventures ever made. Council of Thieves 2-4, Serpent's Skull 4-6, even Jade Regent 2 involve "go to X, retrieve Y." Dungeons work really well given the limitations of published adventures, and, well, you need SOME reason to go down there.
I'd go on, but you get the point. I don't think we have enough data points for an interesting discussion of AP patterns to really emerge. I would further argue that, going forward, we'll find Paizo are far too creative to fall into any sort of rut or pattern worthy of substantial comment (beyond basic, universal plot structure).
