Does a long skill challenge (12 successes needed) feel like a grind to anyone or is there enough roleplay to keep that active?
Even short skill challenges usually feel like a grind to me--they are a fundamentally flawed mechanic. But skill encounters (including that rare, mythical beast, the well constructed skill challenge, if one ever mysteriously manages to claw its way from the well of existential uncertainty) don't have to be grinds, and more importantly, do not become grinds in the same way that combat encounters do. In fact, it is probably not helpful to use the same word to describe them.
A skill challenge becomes tedious not primarily when you are simply repeating the same mechanical actions. In one H3 adventure I played, there was a skill challenge to climb up a cliff. (Many would now argue that this is not a good skill challenge, and maybe it isn't; in any event our DM did not run it using the skill challenge mechanic so it serves as a good example of how a repeated skill roll need not be tedious). Now, in order to succeed on this, our characters needed to succeed on three athletics rolls each. There weren't other applicable skills and though I was playing an eladrin, I figured that the fact that my warlord was good at climbing and that we might be ambushed immediately upon completing the climb meant that I did not want to use fey step to finish the climb. But it still was not tedious--each roll carried a different and easily understood consequence. Failure on the first roll was meaningless; we just made no progress. If we failed by enough to fall on the second roll, however, we were going to take some damage and sufficiently bad failure on the third roll would mean a lot of damage. After the first character got up, he was able to let a rope down and this made it much easier for the following characters until we reached the wizard who was hopeless at athletics. So, we knotted a loop in the rope for her and hauled her up. Was it interesting? Not especially. But it wasn't tedious despite simply being a whole bunch of athletics checks. The defined goal, defined progress to the goal, obvious consequences for failure, and strategic considerations (my warlord and the fighter actually both climbed up at the same time so that we wouldn't end up with one person alone at the top if we were ambushed) prevented it from being the equivalent of a grinding combat.
Instead, a skill encounter becomes tedious when the mechanics of the encounter are detached from the story. This often leaves you repeating the same strategic actions over and over again: think up whatever fanciful excuse is needed to roll my best skill. Even well conceived or (much more commonly) ill-conceived but well-run skill encounters often fall into this category when the skill challenge format is used. If we have reached our destination (snuck into the enemy camp, stolen the secret plans, or whatever) and still have successes left to go, reaching those last successes will be tedious no matter how good the initial portion of the encounter was. At that point, you are no longer rolling dice to accomplish something in the game world. Rather, you are rolling dice because the rules arbitrarily say you need to roll X number of dice before the encounter is over. Tedious.
Likewise, social skill challenges frequently become tedious. The first character makes a good case. Roll diplomacy. OK, one success, but not enough. The second character makes another good but completely unconnected--perhaps even incompatible--argument. Two successes and we're out of things to say. The other characters then proceed to throw increasingly loony role-playing at the wall to support whatever skill they are best in until it mercifully stops. Now, why does it work out that way? Because in general there is no discernable sense of progress or change in the situation. The first player really did most of what would be relevant to the story and the second player did something else that would have resolved the issue in a different story, but those aren't enough successes so the skill challenge is still going. In short, it often works out to be that kind of a failure because the story ran out after one or two checks but the challenge kept going.
On the other hand, a skill encounter can still be interesting even after we have succeeded on a dozen rolls if the rolls continue to be meaningful and connected with the story. Two examples follow:
I'm pretty sure that my most recent DM for Dale 1-7 cut the initial skill challenge short, but in so doing, he managed to keep the investigation phase of the module interesting.
Here's what happened: We arrived at howarts in the sky in the evening, used our thievery skills to rig the elevator so that we would know if anyone used it during the night, talked to a couple of the students and decided to do a bed-check as an opportunity to investigate the student's rooms. Maybe that's a bit out of character for the institution, but our DM ran with it. We walked through all of the bedrooms, noted only one missing student, and found some strings hanging from the windows which we surmised were used by the students to receive their forbidden substances. We investigated the kitchen. Then, we decided to sneak into the missing headmaster's office and found his diary. We went to the spot he had planned to investigate and then followed the trail we found there right to the first combat encounter.
Like I said, I'm pretty sure that our DM ended the skill challenge before we got the requisite number of successes, but whether or not he did so, he ended it at the right time. Once we found what we had found, saying "you have to wait until dawn and teach some classes before proceeding to the next encounter" would have just been rolling dice to roll dice. We had accomplished what the skill challenge was there for. Anything else would be tedious. (And anything less would have been incomplete).
Mini 1-3 provided a good example of the opposite situation: a skill challenge that was over mechanically but needed to be extended further. We needed to sneak into a mining camp. So we waited and watched the guards, then snuck up to the least exposed corner of the compound and climbed over the fence. Then we realized that there were guards at the gate we wanted to go through and that those guards were right outside the barracks where the other guards were sleeping and within a very short distance of the front gate where another patrol was stationed.
Our goal was to sneak into the mine undetected but the skill challenge was mechanically over. Fortunately, our DM was not shackled to the mechanics of skill challenges. We sent a character back out the compound, around to the opposite side (this enabled her to avoid crossing the main walkway where she might have been seen--we weren't sure if the guards had darkvision which would have rendered the shadows no defense against detection), where she climbed over the fence on the other side, snuck into the guard barracks through a window, stole several uniforms, climbed out, came back around, and climbed over again (fortunately, these were all relatively easy tasks). Then we disguised a couple members as guards and proceeded to bluff our way past the guards outside the mine entrance (though we still had to fight those inside--our plan for the bluff wasn't very good).
Now, that encounter could probably have been simplified a bit--the "steal some uniforms" side-quest for the rogue may have been a bit boring for the other players and if it had been accomplished with half the rolls, probably would have been better. But it still is a good example of how an encounter resolved solely through the use of skills and non-combat powers can continue well past the 14 rolls necessary for the most complex skill challenge without becoming tedious due to the amount of dice rolled. And why did it work? Because the entire process was directly connected to concrete goals and situations in the game world and, until it was over, there were still non-combat actions that needed to be resolved.
Now, it is my frequently expressed opinion that the DMG's skill challenge mechanics do not encourage good skill encounters but rather actively interfere with both the creation and the execution of them. That, however, has no bearing on my analysis of what makes a skill encounter interesting or tedious. The bottom line is that skill encounters are interesting when they are connected to the story and become dull when they lose their connection to the story, whether they lose that connection because the story isn't finished but the skill challenge rolls are, the story is finished but the skill challenge rolls aren't, or because the strategies for succeeding mechanically in the skill challenge make no sense in relation to the story.