What doesn't work in a script makes for a fairly small subset of writing strategies. This is why there are so many columns and books about what not to do. The principles of what does work in a filmed story make for a somewhat larger set, one that is expanding as new good stories are filmed, but they are so general (e.g. "the protagonist should have a clear goal") that following them doesn't ensure a story that would appeal to a film audience.

I was a peer reviewer on a now-defunct writer / filmmaker site, and there were handfuls of really well-written scripts that had no major flaws, but I didn't think that even those that made for good reading would have resulted in a successful film (I'm guessing that One Saliva Bubble would have been like this).

On the creative side, I once followed the advice of the original Save the Cat to the letter, and the result was the worst thing I ever wrote. It's not that Snyder's formula was wrong (it was certainly formulaic), but that formulas and paradigms don't guarantee success. It's telling that most of the few screenwriting gurus who have also written commercially successful scripts penned screenplays that would never be used as a model of great writing ("Stop or My Mother Will Shoot").

There is a significant level of magic or art (depending on your metaphysics) in creating the blueprint for a movie, and while great writers have the best intuition about what works, even they sometimes miss the boat. This is how I interpret William Goldman's adage that no one knows anything.