I think that's a very fair critique of Lost. However, just because they did it poorly doesn't mean that it has to be done poorly.
Ok, so we've actually forked out in this thread several related ideas.
a) Cutscene as a narrative hook to engage your audience
b) Scene Framing techniques which are being called cutscenes, even when it isn't technically a cut scene.
c) Cutscene as railroading technique to achieve a story goal
d) Illusionism as railroading technique to achieve a story goal or even set your story goals in the first place.
e) Dramatic pacing issues generally, whether or not they are related to cut scenes.
Let's stick with where you and I started, which is the use of cut scenes to introduce your audience to a mystery in order to get them engaged.
So the first thing I would say is that while you can do this as a cut scene, there is no requirement to do it as a cut scene. You can frame the scene as a participatory scene and simply show the players the scene, and you can use various other techniques besides cut scenes and hand waves to ensure the scene is plausible and works out according to your plot point. Those can be various acts of GM force ("railroading"), clever set ups ("moat between players and action"), or literary techniques ("deus ex machina", etc.). While doing it as a cut scene is very tempting, because its easy and nothing can go wrong, it's a less skillful technique than doing a participatory hook that engages both character and player.
Secondly, if you must dangle your twist or the existence of the twist early, then the skillful writer is doing so as misdirection because they have an even better twist that they want to misdirect the audience from ever even thinking about. The writer assumes that his audience is very savvy and very genera savvy and will know that you have a twist and will be looking for it. The writer knows his payoff is bigger if the audience doesn't figure out the twist until right before the writer is ready to reveal it, so to misdirect the audience the author dangles a lesser twist in order to suppress the audiences keen narrative sense. They'll smugly think that they have the twist figured out early on, and then bang, you hit them with the double twist at the right moment.
There are some good examples of this. If you've seen Sixth Sense, you know that the author pulls a double twist to hide a literary twist that is otherwise very obvious and does so to good effect. More relevant to RPGs, if you've read James S.A. Corey's RPG inspired novel, "Leviathan Wakes", you'll know that the plot actually contains about six twists, and has as a trope the PC continually thinking he's found the right answer, trumpeting out to the universe, "This is the plot"... and continually being wrong. Not only is it a great contrast to an RPG plot being whatever cool thing the PC picks is what the answer is, but layered twists help keep the reader distracted from some of the twists that if the reader figured them out early would diminish the story. This sets up in my opinion a Shining Moment of Awesome.
That's absolutely true. It doesn't change what the characters know, it's purely for the players. It was a way to increase players' interest in the game and the game world in addition to interest players had from playing in the game.
And again, I'm not saying that that is badwrongfun. I am saying that hooking the players but not the characters (or more rarely, the characters but not the players) is an example of lesser skill than doing both at the same time. And as a matter of best practice in our GMing, we ought to be shooting for the moon and not merely what we can get away with.
And the results of the temptation of going the easier route is one everyone in the community ought to be familiar with. We've all had the experience, either of a DM directly or second hand through a published module we've been asked to play by a DM, of a story teller that wants to tell great stories, relying on the well known easy techniques of literary story telling with the result of finding that instead of having been invited to play a game, we've been invited to listen to a budding novelist narrate the story to us.
Nah, we have no trouble not acting on stuff that we as players know but our characters don't.
Lets just save that discussion for a later time.
Part of the backstory of the world is that the old king was killed and in the assault on the castle the king's infant son (Racine) was spirited away along with the royal sword and was never heard from again. One of the players asked if he could be Racine and have the royal sword and the GM said sure. The prelude starts in media res with the current king and queen hiring a bountyhunter to recover the royal sword because they just learned the one they have is a fake. The bountyhunter's payment for recovering the sword is a map to where the elves first landed on this continent.
As a second part to the prelude, in a different part of the world, bandits attack a village and are defeated by a skilled fighter. The fighter vows to hunt down the bandits unless they join his army or turn over a new leaf. When the bandits join him, they ask his name and he tells them it's Racine.
Through the cutscene we learn a little bit about the history of the world. Through the NPCs' dialogue we get a sense of the personalities of the king, queen, bountyhunter, and the Racine imposter and how people react to them. We get a glimpse into conflicts that are on the rise. We learn tidbits like the map of where the elves first landed must be worth a lot if that's the bounty for the royal sword but we have no idea what makes it so special. Now our GM has players who are not only excited to play their character's story but players who are excited about his world and the potential stories that are happening in it.
None of that is necessarily bad, but I think all the plot points you describe can be given to the players through participatory play. The only real advantages to the cut scene technique you describe is that its fast - I can layout plot points in a hurry if you are just sitting there listening to them, and it's easy. That is, it's really easy to create cutscenes compared to creating participatory scenes that develop the story well.
Compare the introduction of a video game like Mass Effect, Half-Life, or Skyrim where the opening of the game invites the player to participate in the game's backstory and games that just show the backstory as a cut scene and then have the player get down to killing rats in the basement. Showing the story as a cut scene is easier. Less is likely to go wrong. Non-interactive story telling techniques are well established. But our goal as designers isn't simply to invent and further the art of literature or movies, but to invent and further the art of gaming. We want to use techniques that can only be done in gaming, because when we do them right, the payoff is so huge. We want to go beyond either showing or telling, into participating.
So while I'm ok with cut scenes, I wonder how much more powerful the game's hook might have been if they were in the village during the bandit attack and were saved by 'Racine', and just how this twist might have hit them at that point. Lots more could go wrong, but I think that plays out as a better scene when it goes right. Likewise, we could have the PC's learn that the King and Queen have announced that the royal sword has been stolen, and that a large bounty has been placed on the head of the thief! Again, that plays out very differently I think if the character is participating in that scene, knowing the sword is on his hip. And so forth. We don't have to do any of these reveals as cut scenes, and by not doing them as cut scenes what we are doing is not only engaging the player interest, but encouraging and validating the player to role play out a response to his character's knowledge.