How we have widened the bottleneck of development

So here's the story. We made 10 board game projects a year when an opportunity came to scale the production and make an order of magnitude more. The problem was with the development: we had to increase the core team no more than twofold. We've managed to make roughly 80 projects. This talk will tell you next parts of the story:

1. How the process was initally organized as kick-in-the-ass-driven development, how we flirted with Agile and other weird stuff and how after that we locked ourselves with a microwave oven and tea and drew on a whole roll of plotter paper.
2. It turned out that a game is not a development process but a building process that looks more like a Boeing logistics. And the most expensive is not to "do on time", but "to do without a mistake on any step".
3. We created requirements from a subprocess to a subprocess, rethought the project manager role, came to the idea "when in doubt, redo".
4. We selected people on different stages, trained them and got them to know our requirements.
5. We separated urgent projects (commercial orders) where the deadline was more important.
6. Model that we got as a result: the process tree, checklists for delivery to internal client, PM role, tracker role, integrating marketing, production and different processes on different levels. A whole new set of mistakes.
7. The result: who gets to kick who, how to solve the paradox "it isn't in production's best interests to produce high quality product", current model and its problems.
8. Communicating with external clients, pain of approval, why it's better economically to put production before approval.