In some ways, the parts I like about D&D 4E get in the way of the things I like about the roleplaying. This is not a simple protest that "you can't roleplay in 4e!" because such a claim is utter crap. I have met precious few games you can't roleplay in with the right crowd and inspiration. No, rather, the things I enjoy about roleplaying are not the things that draw me to 4E. I am, not to put to fine a point on it, drawn to the shiny bits. I want to make characters, create power combos or team-ups, and get in fights. And then, more problematically, I want to do it again.
This means that every character has a fairly limited shelf life. I want to try out certain powers and build ideas and then, once I've seen them in action and am satisfied, I want to try something else. 4E has built their castle on a foundation of unending novelty, and while I could attempt to resist it and make 4E something it's not, I can't help but feel that I would be better served embracing it.
Taken to its extreme, this logic would suggest that I want nothing but fight scenes. I actually admit that's not far off, but it fails to capture the whole picture. I absolutely want advancement and the sense of an arc that comes out of campaign play, so the question I face is how to mesh these things, and I propose that a solution might be radically changing the campaign structure.
Blake Snyder has a neat model for planning out a movie where he spreads out index cards in three rows to represent the story beats of a movie. He's got a particular formula that works or movies, but I'm mostly looking to steal the structure, three rows of index cards, each one noting a scene (in the case of 4E, a fight scene or possibly a Skill Challenge), as a way to handle a campaign. The idea is to think of a campaign as a limited number of set pieces (i.e. fights) to be run through, with advancement and a small amount of connective tissue between the scenes. It is, to be frank, quite blatant railroading, but provided that there is no confusion or deception on that point, this is not necessarily a bad thing.
Index cards in rows ends up being a good model for this for a few reasons. First, the rows constitute a rough "act" structure, so you know to throw some of the meanest stuff into the second act (the second row) so the third act (row) can be the climb to victory. Second, the ability to spread it all out to look at can help inspire ideas. Last, it's cheap and convenient - fancy mapping software may be pricey, but a stack of index cards and a pen are within most budgets.
The trick of this is that you are basically discarding existing framing mechanisms (dungeons, overland maps and encounter tables) and replacing it with something more obviously structural, tying the scenes together much the same way they might be tied together in a movie or TV show. Practically, this is pretty easy to implement. Let's say I want to do a 9 fight heroic campaign; that's 3 rows of 3. In term sof a narrative I want the heroes to face a local danger, be imperiled revealing the greater danger, then face the real danger. That's 3 acts/rows right there.
Act 1 is all about fighting off raiders and tracking them back to their lair. Figure 1 scene with raiders attacking town, one scene of pursuing the raiders because they've got hostages, one scene at the lair.
Act 2 reveals the danger as the heroes discover the raiders are just the tip of a bigger iceberg. Maybe the big bad has an army of undead under construction or something. Scene 4 is in the caves beneath the bandit's lair, fighting some guards. Scene 5 is a skill challenge as they sneak around the secret base. Scene 6 is them getting captured and thrown into the thresher pit and fighting certain doom, maybe with the help of an ally)
Act 3 is the escape and revelation that the master villain is the local lord, and a showdown. Scene 7 is the escape up secret tunnels into his castle and a fight against his guards. Scene 8 is a fight against some of his now-active undead army. Scene 9 is a struggled on top of the dam above the valley ending in wiping out the army.
Now, that's pretty rough, but if I spread these out I can probably refine these scenes until they work, looking to future scenes for ideas about previous ones. I also think about how much of a level range I want to showcase, and decide I'll do level and gear bumps after scene's 3 and 6, jumping to levels 4 and 8 respectively. Beyond that I have to flesh out some details about the lord, the valley and stuff, but not much, or more specifically, no more than I'm interested in. How much happens between fight scenes is really a function of a groups interest in that part of the game. If your group wants a lot of non-fight material, that can be handled just as easily as if they want to just jump from fight to fight, connected with a little bit of narration. The only thing impacted is how much time it'll take.
And with all that, I've just sketched out something that will be played to a satisfying conclusion over the course of 9 fights, perhaps 3-6 sessions of play and which, to be frank, is not going to look too different than the net result of playing through most published adventures, except this was faster and easier to plan.  With a little time and familiarity with the model, it's easy to introduce branching or flexible scenes that change based on player choices, but that's the easy part. The hard part is the serious shifting of gears this model calls for.
It's not going to be to everyone's taste, but for a group comfortable with the abstraction, it becomes a faster way to engage the material that excites them. For me, the ability to finish a story, test out a build, and move onto the next thing over the course of a few nights (rather than a few months) is hugely appealing.
1- It is incredibly important to note that this is *a* solution. It is merely a proposal for a way to structure a 4E campaign, not a proposal that all 4E campaigns should be done this way.
2- Tangentially, the one rub of "beat" terminology is that it's actually used for several things, including scene beats, story beats, and beats as in a moment of time, such as for a pause.
3- Yes, this thinking works very well with that idea of positive and negative charges to scenes from yesterday.
4 - This speaks to one of the likely objections to this approach. Players exploring the dungeon and engaging with it as a place is an essential part of D&D, and this removes that. To this I say, 4E already removed that, this just acknowledges that fact. I don't say this as a criticism, but the simple fact is 4E's relationship with the dungeon is different than previous editions. I actually kind of like that, but it's jarring if your expectations were set by previous editions.