The Skill Challenge framework, as a narrative-sculpting pacing mechanism, answers each of those questions tangibly. It doesn't just hint at a framework that vaguely alludes to a formula. It is a formula.
This is the beauty of the Skill Challenge framework. It is also the tragedy of the examples in the DMG.
The examples in the DMG are just that examples. However, some people took them to be literal examples. For those people those examples became the "only" way to use the framework. The "arguments" about the "intimidate" check on the duke challenge were notoriously "literal".
It was very disappointing. The use of the framework, in published form, became incredibly mechanical. Each subsequent "publication", by its mechanical display, reinforced the erroneous view that it was all about the mechanics. When in reality what should have been pushed forward was the use of the skill challenge to do scene framing.
If the section on the book had used "narration" at the game table to show how a skill challenge was meant to be run, it's possible that all the "confusion" might have been reduced.
When I design skill challenges the entirety of it is much more organic. I also don't have to write down mechanically what happens. Very simple notes work fine for me.
When reading a published skill challenge what I do is reverse the skill/outcome sections and add complications. IMO, what is really missing from the framework is a "failure escalation/complications" section. I also prefer the variable failure method rather than the static 3 failure method.
If I was to write a template for the skill challenge framework I would change it to something like this:
Skill Challenge
Complexity: X (Y successes before Z failures)
Description / Goal:
Method
Action description / relevant skill / complication / difficulty variation
Success beyond expectation: Goal accomplished + bonus
Success: Goal accomplished
Partial Success: Goal partially accomplished + penalty
Failure: Failed goal
Catastrophic Failure: Failed condition + added penalty
I would also have taken more time in explaining how to design an exciting skill challenge. A skill challenge is not the same as a skill check, and that needs to be greatly emphasized. If a simple skill check is all that is needed then don't design a skill challenge around it. If a scene is what is needed then a skill challenge might be an appropriate tool to use.
I find that the framework works much better when it's used to make "freeform" cyclical OODA (Observe, Orient, Decide, Act) decisions. The DM frames the situation (GOAL), the players observe and act (action description), the DM determines the appropriate skill to use (relevant skill & difficulty), the result of the action either advances their goal, or adds a complication (complication description). In my games this is all descriptive, even though skill checks are used.
It's also tragic that some decided that this was simply an exercise in dice rolling, instead of what is actually explained in the book, a descriptive form of extended task resolution with relevant consequences.
When used in the appropriate manner I can pretty much turn any exciting scene into an appropriately framed skill challenge:
Searching for a bandit in a city.
Helping a family/town save a burning farmhouse/building.
Traversing a mountain range/tundra/jungle/etc.
The Indiana Jones mine cart chase.
Getting to Theoden, through his "guards", and freeing his mind from Saruman's influence.
Rallying the demoralized Theoden to go out and face the orc's at Helm's Deep.
Convincing Theoden to send troops to help Gondor.
Disarming the doomsday device beneath parliament while goons are protecting it.
Crafting of an artifact.
Infiltration into a "masqued" ball.
Many more.
[Edit]
Forgot to add Partial Success
-