I'm using a lot of variations on the basic skill challenge idea, and I like feeling like I can be very flexible about how to run them -- the core mechanic is a starting place, a frame on which I can imagine each new challenge.
At the same time, I respectfully disagree in many (not all) cases with the idea that allowing the players to see the scaffolding of the skill challenge structure is a bad thing -- but in the end I think that there's a wide variety of groups that play, and a wide variety of tastes, and ultimately there isn't a "right" way to handle that.
For me, I don't see the point to having such a structured system if that structure isn't going to be made accessible to the players in some way. If you're going to hide it from the players, why use the structure in the first place? On that path, the only thing that has been created is a formalized structure for the sort of thing a DM would be doing by imagination and the seat of his pants in previous additions.
At the same time, showing the structure to the players does draw attention to the game elements at the expense of the role playing elements. I'm willing to make that sacrifice for the advantages I feel I get out of bringing game elements into those moments at the table -- but clearly that doesn't work for everyone. It also doesn't honestly work in all situations.
I see three different broad categories in my own game . . . . social, action, and investigative challenges.
For social challenges, unless I'm playing with new players, I'll frequently not make obvious the structure of the skill challenge system. I may use some of my tools for indicating successes (I throw poker chips on the table, green ones for successes, red ones for failures) as we talk and I ask them to make skill checks. I do, however, try to take a moment in the flow of conversation between me and the talkative players to ask what the other players are going to do while this is going on -- to try to give everyone a chance to get involved in the challenge one way or the other -- without forcing the issue. But, I show the least structure in a RP challenge because those are challenges that seem to work best that way -- it was a part of the game that we already did pretty well before this edition. I also really like the idea of using
degrees of success in social encounters rather than just having a succeed/fail result.
For Action encounters, I tend to show the structure much more -- whether it's finding their way through the wilderness, trying to help a town fight a fire, trying to chase down a ship at sea, or any other variation on these sorts of challenges, the action challenges fit much better into my game with strong structure and game element. These types of encounters, IMO, make a much better parallel to combat encounters, and I like being able to give them the same sort of structure as a combat encounter. At the same time, I don't think that's the right choice for ALL of these encounters. I'll usually start an action encounter by explaining that the PCs are going into a skill challenge, what the primary skills are, and an encouragement to try other skills if they can figure out clever ideas. What this does is help me flesh out the details of the scene with the players, and help them see the ways they can interact with the world.
For Investigative Challenges, I ask myself a question -- can the game go on if the PCs fail the challenge. If the PCs can fail the challenge and the game still goes on, I'll run a skill challenge for them. But if they really can't reasonably fail the challenge and continue on the path they're on, I don't use the skill challenge rules. I have been using something similar, where I have a bunch of information the PC can gather with successful skill checks in a variety of areas, but I don't bother tracking failures -- they're going to keep making checks until they get bored with the investigation or until I run out of information to give them. I could just dump the information out for them, but by making them make skill checks to earn the information, and giving each player a nugget of information that they've earned with that check, I keep them engaged in gathering that information and with the story (my players tend to grow inattentive if they're not swinging swords).
So, YMMV, but for me I think it doesn't make a lot of sense to make blanket statements about skill challenges -- how best to run them. I think it depends in large part upon the specific challenge in question and the group you play with.
-j