Not every entry in this series is problematic or controversial. In fact this was one of the best examples of team building I experienced in my process. The idea for branching storylines was practically a given when I first started with concepts and it was only a question of how far I wanted to take it to the extreme.
I found that some concepts are far to wild to run with and the cutting floor is much harder than an early cut. I originally had an idea for using D&D style decision trees where during dialog you could use your triggers and shoulder buttons to select an alignment then your face buttons for the choices. The team jumped on it with new sub teams for each alignment generating an overwhelming amount of content.
Then came time to make it a reality with voice actors and game scripting to make it all work. Sadly we found almost immediately that the amount of dialog and options was going to be massive and yet we kept trying to find ways to make it more efficient and keep as much as we could. In testing we found even with visuals on screen to help players most of them picked the same options over and over leaving plenty of content unheard. When we asked them why it sounded like the quality of the writing was uneven so certain choices were just inherently better. Even with revamps though it turned out some options were just out of place in a given scenario.
I then questioned what kind of a game I was really making. I knew from the beginning it was going to be a somewhat linear storyline and the primary point of the dialogue tree was to show that choice was really just a facade and that some things are meant to happen or are unavoidable. At the same time I didn't want just a linear style action game especially with all the other content we'd already started to produce.
Then came the hard choice of deciding how to change the decision trees and what scripts to cut. I had our testers pick which options they liked most and narrowed down our choices to a bland single page multichoice system more in the style of quantic dream games (beyond two souls, heavy rain) than bioware (mass effect, dragon age) and eliminated the morality of it all completely stripping out alignment. In some cases we completely removed choice simply because our testers unanimously agreed on a single option and thus we made it an automatic response.
I think everyone understood when I announced the cut that it really was necessary and while the implementation was far from perfect it wasn't the actual problem it was more of an identity issue of what the game really was and how the system itself was out of place. My future recommendation is that if you're going to include dialogue trees, seriously consider them in advance and don't do voice acting till you've already tested everything else.