I remember back in 2010, I had just written my first part of a setting (Akos). I had all those ideas bursting in my head: new character options, new races, complex political interaction between NPCs, between nations and between factions. It was a grandiose experience and I was so happy about my work. The draft was rough, sloppy and needed cleanup and tightening, but it was ready.
I happened to fly to Montreal in the weeks that followed and decided to put a game together for my friends. In advance, I sent them the document with a few goodies and solicited feedback.
It was honest (I'll stay just shy of "brutal"). And I learned a number of things from the experience. Things I brought into my later projects and finally into Tyrants of Saggakar.
One of the lessons a friend of mine (Alex-FG) phrased so eloquently "You need to have at least one story or adventure idea per page."
I was taken aback by the simplicity of the design rule. I like simple, elegant design rules simply because they are easy to follow and adapt. Overly complex ones eventually forces you to adapt to the rule rather than letting the rule be a guideline.
With time, I have evolved this rule to mean "every page, must present a conflict or a situation where adventurers can come in and explore." A mouthful, but in essence, everything written must lead to something the PCs may need to investigate. The less secrets are defined, the fewer prophecies explained, the better. If you think Joe killed Bob, leave clues, but do not actually say it. Revealed secrets take away from the GM's ability to write or create this adventure per page.
For me, once something appears in an adventure or in a book, that is set in stone. I don't go ret-conning or changing things because I have a new, better idea. So with this prudent approach, I define the minimum and work up from there. In many ways, this is what I see in Paizo's Golarion. Bare bones really, then things get defined up from there.
This rule evolved into my adventure writing philosophy. Focus on the important stuff, that which is required for the adventure, and move everything that is not out of the adventure. This gives the GM freedom to work within his particular style, but also limits the amount he needs to prep and remember an adventure. By giving him the tools he needs to handle the situation: what happened before, what is happening now, who is involved. However the PCs evolve from point A to point B will be up to them. How one group will do it depend heavily on their style, abilities and interests. Some will try to brute force, others sneak around, others use diversions, yet others will use the charismatic way through the problems.
The important thing in adventure design is not to burden the GM with a whole lot of information that may not be relevant to his PCs' evolution through the adventure. Nothing worse than watching "GM-loading" as he frantically searches a document of an answer. Focus on the essentials.