It's not small in itself.
But over a time of 11 months of development, and considering we are talking about costs, it represents a pretty small investment.
Compared to AAA title production costs, definitely.
If you spend less on graphic content, you can spend more on writing. Instead of building your team with say, 10 graphic artists and 2 writers, you can hire 6 graphic artists and 6 writers.
Assuming you're paying the same salary. And assuming those 6 writers are all more or less at the same level and you have someone in charge of coordinating their efforts... Otherwise you have a very uneven game. Who knows, maybe the 2 writers from your example are able to come up with better dialogues and stories, but charge so much the studio can only afford 2?
The Torchlight example was given to illustrate my second point, which is that today, with the range of production-level open-source components available, you can drastically reduce the investments in time and money required to make a game.
Again, "it's not that simple"; having an open-source engine is one thing, building the 3D assets, UI and characters you need is another cup of tea. You can reduce your costs, but the bulk of the work is by no means "already done", as you seemed to imply in your first post. You need to build levels, the aforementioned assets and actors, make all the animations and integrate the lot into your story, while making the whole thing stick together in terms of performance.
Which is especially valid regarding the RPG genre since they don't demand high-tech state-of-the-art engines.
True. Nevertheless, although RPG's don't usually require the level of eye-candy the FPS genre is determined by, they're usually done over a longer time period, precisely due location amount, quests, writing.
And costs are mainly determined by your team size, team experience/skill and time of production.
