As I see it, we've got three major issues that are most guilty of threatening the meaningful decision in games. These are also examples of problems which would naturally be avoided if game designers adopted my philosophy for games. They are character growth, saved games, and a story-based structure.
Character Growth. Ideally, a game should be increasing in difficulty as a game progresses. However, we've now got an expectation that our character -- our avatar -- will gradually increase in power as the game progresses.
Of course, designers try to make up for this by cranking the late game difficulty further, but this is a very bad position to put yourself in, and it's one of the reasons why we in video games have such trouble balancing our games.
Essentially, you're trying to hit a moving target. Assuming that the player can become better at the skill of the game, and the character can also become more powerful, and both of these can happen at somewhat irregular rates, the prospect of balancing late-game difficulty becomes impossible.
Anyone who's played a Final Fantasy game through to the end can back me up on this (I remember the final boss of Final Fantasy VII being pathetically easy for my Cloud to take down.) I think the designers of such games are aware of this issue and prefer to err on the side of "too easy".
Of course, if your game is too easy, then your decisions are no longer meaningful (as my decisions weren't meaningful in my Sephiroth battle -- I think it was a foregone conclusion just based on character stats alone).
Saved Games. I call the quicksave/quickload (or any similar system) "the most powerful weapon ever wielded" in a video game. This one is so straightforward that I can keep it short: essentially, a player's job is to try to play his best; to try to make optimal moves. The game allows you to save and load whenever you want. So, when faced with a difficult decision, what is the logical thing to do? Save the game, then make the decision. Well, looks like that was a bad idea! Re-load the game, and try Door Number Two. Hooray! I'm so good at this game!
The issue with saved games is that they insulate us from ever having to make a meaningful decision, a decision that has effects on the game. If you can reload after making a bad choice, then that choice gets no chance to have effects on the game. If you can save the game right before every challenge, then it is no longer a contest. Once again, it's a foregone conclusion. It's only a matter of when you win, not if.
I should mention that there's a common counter-argument to this argument that goes something like, "well, if you don't like to reload the game after messing up, don't do it". The issue with this is that I am having to create extra rules, "house rules", if you will. I am having to do part of the game designer's job, and that isn't fair. Furthermore, many games are actually balanced with this in mind as an element of gameplay, and due to my next item, it's actually rather unavoidable...
Story-Based Structure. Never before video games was there this idea that games get "completed". Instead, games were played in "a match". Now, all games are expected to have a long campaign, capped off by a credits reel. This completion-based mindset has dire effects on our friend, the Meaningful Decision.
Firstly, most story-based games are quite long, with regards to games from throughout history. While most games historically have taken between ten minutes and a couple hours to finish a match, modern video games aren't considered "finished" in any sense of the word for twenty or more hours.
This on its own isn't a problem, but it also means that it becomes a bit cruel and harsh to actually ever give a player a meaningful "loss" condition. So, that means all that they can do is win -- therefore the meaningfulness of their decisions is destroyed. All they can do is beat the game slower or faster; it's no longer a competition.