It would be interesting to know if working for Obsidian and Inxile is just as bad as working for Crytek. Many people, including myself, had high hopes for games such as Pillars of Eternity and Wasteland 2. That didn't work very well. Is the process to blame for those games' mediocre writing? Or did they just choose the wrong people to be in charge of the narrative?
I have no personal experience with and don't personally know anyone who worked at mid-size A to AA studios, such as inXile and Obsidian. They're not full-on AAA but they're not small either. I'd assume that their workflow is closer to AAA than to indie teams with 10 people though.
When I interviewed Swen before the release of Divinity Original Sin 2, he said there's about thirty writers working on the game. It's not an exceptionally huge game though - what does it need THIRTY writers for? The writers were even working from different locations as Larian had several studios at that point: one in Ghent, one in Dublin, one in St. Petersburg (IIRC). With so many people in so many locations, coordination becomes harder than when you work with a small team that lives close together.
When I worked on Realms Beyond, that was even before the successful Kickstarter so there were fewer people on the team, maybe 10 at max. We coordinated online and regularly talked about setting details, quest brainstorming, etc. Even with a small team you had to do lots of communication since different people had to coordinate their efforts: when you write a quest, you have to coordinate with the level designer who builds the town your quest is set in, and you both work off each other - you give instructions to the level designer on what you need for your quest (I need a hunting lodge for the local huntress and she has to have a dog at her side), the level designer gives you some ideas on what to do with your quest (I placed a little toolshed next to her hunting lodge, there's some items in there that might come in useful in your quest idea, maybe you can add the option of letting the player steal them), etc etc. This kind of coordination works very well in a small team, everyone can work off each other and there's a definite personal note in each quest, after some time you will easily be able to identify who made which quest and which location because everyone has their own style of quest and area design.
Now imagine the same with three, four or even up to ten times as many people. Coordinating like that is going to be much harder. You're probably not going to talk directly to all of your colleagues, instead you'll just talk to the lead guy and he will consolidate all the suggestions and give people the okay or tell them to change their ideas. You end up with internal inconsistencies much more easily, and there's a high likelihood that the game story will feel disjointed rather than every quest harmonizing with each other and with the locations it's set in. Someone might be working on a dungeon you don't even know exist, and the questgiver is a character living in the same building you are currently designing... but you don't know this NPC is going to be placed there. With a larger team, it's impossible to know what everyone else is doing. When you have a team of 10 or 20 people in total, you are always aware of what the others are working on.
Ah yes, John is currently building a cave network... Lydia is writing the characters of the noble house of Greywinter... I have an idea for a quest involving an old artefact once owned by that family, and it got lost in the ancient caves... gonna talk to John and Lydia and throw some ideas at them, maybe we can work on a quest together!
Meanwhile in a big team of 100+ people you're more likely to think: John? Lydia? Who are these people? Have I met them before? Are they new hires or were they always here? I don't remember. Eh whatever, they're probably working on areas of the game far away from the one I'm working on so it's none of my business!
When Obsidian and inXile had big name writers like Chris Avellone or Patrick Rothfuss work on their games, they didn't have them design a consistent area, or questline, or whatever - they had them design companion characters. Those are usually rather disconnected from everything else in the game (apart from the mainquest, to which they will occasionally react). While companions will often comment on things and have their own personal companion quest, they don't have to be integrated as closely with the world as a lengthy questline would have to be, or a location. You don't have to coordinate with level designers as closely as you would if you were to write a questline, because once a companion is recruited he is part of your party, no longer part of the world around you. While designing a questline often means level designers have to build new dungeons according to your design, designing a companion means you just have to look at the finished main quest outline and have the companion react to the characters and places the player comes into contact with.
This seems to be a general trend in bigger companies with lots of writers and designers on the team. You don't coordinate as much, instead you work on snippets that can stand on their own to some degree because that reduces the amount of required coordination (because coordinating your creative work with other team members does take a lot of time).
In the end, large teams means you end up with many cooks who aren't fully aware of what the other cooks are doing, all tossing their own spices and ingredients into the same bowl.
