As hsi379 has pointed out, the way to actually incentivize everyone to participate in this mechanic is to give everyone a roll per round, and then limit the numbers of rounds and base the number of successes required on that dynamic. That way you haven't made everyone else worse off if your check fails. But that, I fear, becomes unwieldy quickly. Either the fiction lurches frequently to keep things interesting for 12 rolls to accomplish one objective, or you're having a lot of folks just roll the same thing to do the same thing lots of times over with cosmetic changes in the fiction if any.
The more I chase my tail on this, the more I think your 1. is a "good enough" solution that's actually hard to improve upon: 3/4/5 for easy/medium/hard, + intro tengwars for councils, that requires 3/6/9 successes, and the LM just has to have a twist or two in there to keep it interesting.
I don't think it's always hard to come up with 3 rounds of interesting, although probably Councils are the hardest version. That said, there are times for shorter challenges as well.
I think I'm leaning toward 1, 2, or 3 rounds with one roll for each PC as the "dramatic heft" portion of the equation.
The length will not have anything to do with difficulty, just "spotlight" or level of importance in terms of table time.
Then
the # of successes required in 1, 2 or 3 rounds sets the difficulty. You make all the tables with probabilities for all the combinations in advance and as the LM you will know roughly how hard a challenge you are actually setting up.
As you mentioned, this has the advantage that a PC is always better off trying to contribute than sitting it out, even if they aren’t super skilled in that area (social, exploration, etc.). Which is kind of when I want to use these skill challenges – when everyone should be involved. (You can always do a shorter challenge if an Expert is doing all the rolls).
It doesn't address the "game" element, but it creates a coherent system.
So for example, let's take the boat example for Skill Endeavors (p.132) in the book (which as is I think is a horrible example) and add some parameters to make it a better skill challenge -- -you know from some friendly horse riders that an orc army is heading your way and will likely catch up with you by sunset at your current paces. You found an old boat and want to repair it and carry it up a hill to the river to escape. If you succeed, you can escape on the boat, if you fail you will have to try to hide or likely be captured (another skill endeavor perhaps!).
The stakes are pretty high and you want to make this into a dramatic scene. You could do a 1 round Skill Endeavour, but you decide 2 rounds is best as part of it will be fixing the boat and part figuring out a way to carry this heavy thing up the hill.
The party is 4 PCs with average skills around 3S. You figure the boat is pretty old and this will be a tough challenge for PCs of this ability and failure is interesting as well so you set the challenge at 9 success in 8 rounds which says around 44% success rate at TN15 without spending resources . [Note: you will have the tables so if you decided on 1 round or 3 rounds you could have also looked up the # of success to equal something like 44% success rate if you wanted. The number of rounds/checks is just for narrative weight. 8 successes in 8 rounds = 59%, 10 succ in 8 rounds = 31%, 5 succ in 4 rounds = 40%, 4 succ in 4 rounds = 60%).
Example:
2 round skill endeavor representing 4 hours of work.
1st round – repairing the boat.
I use
Explore to see if suitable vines or roots in the area. Uses Hope. Success + Tengwar = you find just the right kind of roots and vines.
Athletics to harvest the vines in time. Success. You manage to harvest enough vines and roots to allow for patching the boat
Hunting to thatch the boat together with the vines. Success. You manage to patch the boat adequately enough that you think it will float.
Craft to carve an oar. Failure. You carve an oar but the wood is brittle and not likely to last.
4 successes needed out of 9.
2nd round -- getting it to the riverbank.
You could have free form like the above, or you could decide there really is no getting around physically carrying the boat up the hill. So you say:
Everyone has to make an athletics check.
The low Body hobbit says he would be fairly useless carrying the boat, so he suggests that he will help carry but his primary contribution with be a Song about how Bullroarer Took once carried am entire oak tree from Hobbit to Bree for replanting. This is great, so the LM allows.
So 3 athletics checks rolled all at once then Song.
All 4 PCs spend Hope on these checks. 3 successes on Athletics and 2 success on Song.
So, LM describes how they toil to move the boat and almost give up in exhaustion but the song of old Bullroarer inspires them on for the last push they need to get the boat to the river bank.
You will have to do the work of making up a fairly extensive charts of probability, but once done can just be consulted so you can match the fictional positioning to roughly the probability for success that makes sense.
Note this is still a poor "game" in terms of tactical choices and the actual skills used really contribute more to the color of the scene than anything else BUT it fulfills the role of:
1) gets everyone involved
2) makes a task matter more by giving it more screen time in terms of 4, 8, or 12 rolls and decribed actions (w/ a party of 4)
3) allows the LM set approximate difficulty of a multi roll challenge to fit the fiction
4) resolution in non-arbitrary
Now that I look back on it, I don't love this example I gave since it doesn't show the LM altering circumstances in reaction to success and failure rolls but I'll keep it in because it is certainly one form of using it.