The most rational approach to a skill challenge was to have the player with the most relevant skill make all rolls.
<snip>
Forcing a character to participate against the wishes (and rational analysis) of the player is a kind of railroading
This is not the first time I have seen comments along these lines about skill challenges.
To me, the analogue to this sort of comment in relation to combat would be "The most rational approach to a combat is to have the player of the fighter, who has the best to hit and the best damage, make all rolls - and forcing another player to participate would be a kind of railroading". But in fact I've very rarely seen combat run in this sort of fashion in D&D (certain arena combats would be the exception) - generally the GM frames the combat such that all the PCs are implicated, and so all have to make rolls, or otherwise declare actions, even the unarmoured, staff-wielding mage who is all out of fireballs for the day. And I've never seen that labelled as railroading - it's just playing the game, and the GM doing his/her job of dragging the PCs into the action.
In a skill challenge, the dynamic needs to be the same. This means that "You see an [chest/mountain/whatever] that needs conquering - what do you do?" is no good. Because the obvious answer is "The thief unlocks/climbs/otherwise deals with it." Rather, if the GM wants the player of the fighter, or wizard, or whatever to get involved, s/he has to frame that PC into the challenge. For instance, "The duke turns to Derrik, the dwarf fighter, and asks "So, Lord Derrik, what's your opinion on the matter?" - and now Derrik's player either has to make a Diplomacy roll, or come up with some other cunning plan to get out of answering, or look like a fool, or whatever - or maybe another player spends an interrupt resource to make a Bluff roll instead and draw the duke's attention of Lord Derrik onto something else.
That's not railroading, that's just the GM putting the players under pressure by framing their PCs into challenges - which is a pretty mainstream way of running an RPG.
Seems to me that the problem with skill challenges is that the 4e DMG presented them as being like a combat system for noncombat scenes, when they're not analogous to the combat system at all. Completely different kinds of fun. Skill challenges are not a *game*; if you approach them as a game then the roleplaying will feel "tacked on" because it doesn't affect the underlying mechanics and doesn't affect your chance of winning or losing. Skill challenges provide a structure that encourages everyone to contribute to the narration of noncombat scenes and makes it difficult for the DM to railroad them.
I think that skill challenges are a game to the extent that you can win or lose in them; and you have resources to spend on them (eg skill buffs, action points, and the somewhat underdefined "advantages" introduced in Essentials as a mechanical device to help deal with the rather wild skill bonus disparities of 4e PCs). But I agree that the narration is key - you can't make a skill roll until you narrate your PC into it, and establish (either explicitly or, more often at my table, implicitly) a goal for that roll that fits within the etablished and unfolding fiction.
And I agree it makes it hard for the GM to railroad, because the GM can't bring things to a close until the challenge is ended, nor declare that things are still unresolved once the challenge is ended.
They should only be used with players who like to roleplay for its own sake or for the impact that their description has on the narrative. They should not be used with players who only like to describe their skill check in the fiction if they can squeeze some sort of tactical advantage out of it.
I agree with the second sentence: you need to describe your skill check beyond simply the eking out of tactical advantages, because without that noone knows what is going on. In your first sentence, I think I agree with the second clause more than the first, because the first sort of player I think might find the mechanical aspect of the skill challenge to be a distraction.
In my game on Sunday, the PCs entered Mal Arundak, a bastion established by the god Pelor on the Abyss to guard an artefact (the Crystal of Ebon Flame) that is too corrupting to be stored anywhere else (see The Plane Below for the full story). The bastion is guarded by angels, but they have spent so long on the Abyss with the Crystal that they have been corrupted. In the description of this place in the The Plane Below book, there is backstory, and then a description of how things will turn out if the PCs go there (eventually the corrupted angels, who don't recognise their own corruption, will become paranoid and turn on the PCs blah blah blah) but no account of how to actually resolve the interaction between corrupted angels and PCs. I resolved it a skill challenge. The PCs were very courteous to the angels (multiple successful Diplomacy checks) and eventually proposed a collective prayer to Pelor whereby they would all reaffirm their commitment to defending the bastion in Pelor's name. The invoker-wizard deva PC then used his Memory of a Thousand Lifetimes to recall a teleport sigil in Pelor's palace on the outer plane of Hestavar, used a Planar Portal to open a portal into the palace through which some of Pelor's radiance would shine, and then used a Religion check to bathe the corrupted angels in that radiance, with the aim of cleansing them. The Religion check was a success (between stat, training, feat, item and an Epic Destiny feature it's very hard for this PC to fail Religion checks) and so the angels were cleansed, and able (in retrospect) to realise the corruption that had befallen them.
For me this is illustrative of how skill challenges work - the players have to actually engage the fiction via their PCs in order to establish the ingame possiblity of realising their goals, and the outcome is determined by the mechancial resolution of the PCs' attempted actions rather then predetermined either by the players or the GM. It also shows that, in 4e at least (and I think here 4e resembles "free descriptor" RPGs like HeroWars/Quest, Marvel Heroic RP etc), the game depends upon the players and GM reaching a shared understanding of what is feasible that is not dictated by the mechanics, but is rather a precondition of deploying the mechanics. For instance, nothing in the 4e mechanics states that "the DC for using Religion to channel Pelor's radiance so as to cleanse corrupted angels is X". The idea of cleansing the angels in this way was the players'. Given that he's an epic-level deva invoker it seemed plausible to me, but I (as GM) made him use Memory of 1000 Lifetimes (an encounter power) in order to recollect a teleport sigil in Hestavar (and that means the ability is not available for use later in the encounter). If you don't have players who are happy to improvise in this sort of way, at the interface between fiction and mechanics, I don't think skill challenges will work so well, as you won't see so much dramatic or creative stuff taking place.
