Having a game that makes choices matter is harder than it sounds. As it carries the consequences of having to develop the story for every choice consequence, right down to all the assets sometimes. For example say you have your hero make a choice to stay the night in a village or press on to plunder some forgotten crypt. The consequence being that if the hero decides to plunder the crypt, he won't be in town to stop the orc raid from destroying it. Now you have to generate 2 towns. 1 thanking him for his help stopping the orcs, and the 2nd a destroyed town with maybe one or two survivors to tell the story. Making the hero miss out on a valuable place to stock up on supplies.
That's just a simple example with not really any far reaching consequences in it. Maybe after saving the village from the orcs the hero goes to the crypt anyway. While in the crypt he has to make another choice between setting free a Lich to help kill a master Vampire, or joining sides with the Vampire to kill the Lich. Both of which will have drastic consequences for the nearby village. For the developer as well as now you start having many adventure states for this one village. You have destroyed by orcs, Saved from Orcs no control, Under Vampire control, and under Lich control. It just gets even worse from there, if you stop to think about what happens if he doesn't stay in the village to save from the orcs, and still has to make that terrible decision between the Lich and Vampire.
That's just a simple example too. The vastness of how our choices impact the world around us is mind blowing yet we in real life do it every day. It's even larger when you try to duplicate it in a artificial world where you have to construct every little thing. How many versions of this one village is a developer going to want to reasonably make before they really want to move on with the story. How many varying stories do they wish to include in one game before it's time to 'create a new game.' There's a very big reason Triple AAA titles boast only 10 to 20 endings. Or only have very select choices actually matter in play. The simple act of a single choice in a game can quickly spiral into a vast loop of cause and effect that becomes far too big for the companies that have the time, money, and backing to do them. Let alone some singular dude with a half crazed idea, some extra time, and a bit of motivation to put that idea down for others to enjoy.