2
I agree with nearly everything in this post, save one point: that every encounter in your adventure has to further the plot of the adventure.
I’ll clarify my position by saying that every encounter should have a specific purpose, but I don’t think that that purpose must be attached to the current plot. After all, if every encounter has something to do with what’s currently on the to-do list, you run the risk of making it seem like the entire world revolves around the PCs (which it does, but it shouldn’t seem like it). Sometimes it’s good to pepper your adventures with seemingly random encounters in order to add verisimilitude to your game world; sometimes, in a dangerous fantasy world, the owlbear is just hungry.
But, as I said, every encounter should have a purpose. The lion’s share should be tied to the current plot, and should be furthering it in some way. A few, though–probably no more that two or three in an adventure with 15 encounters–should not. They can be there to add color to the world, to introduce an enemy faction that you plan to use later, or they could be a form of the spaghetti method: throw a few different encounters at the PCs and see which one “sticks”; that is, which one do they latch on to the most? That’s a plot hook for future use.
I’ll clarify one further point: when I use the term ‘encounter’, I don’t mean ‘fight’. In D&D, there’s a tendency, I think, to treat every encounter as a fight, but it’s often more satisfying to vary things somewhat. Social encounters are encounters, too, as are periods of investigation or even research, and even long-distance travel through dangerous terrain, like a desert or mountain range, can be handled as an encounter in 4E. Also, if all 15 of those encounters are fights, it’s going to take you a long time to get through your adventure. Social encounters, travel encounters, and other non-combat encounters tend to be quicker to run, and can be used to build tension and world color just as effectively–if not, in some cases, more so–than combat encounters.
