Seriously doubt that. I am has tied in as anyone, and I am 100% confident that there are projects I don't know about. Heck, every couple of weeks, I'll hear about one that has been going on for months that I was unaware of.
First off, the R&D costs could easily out weigh any cost optimization, and can also impact time to market. You are also talking about R&D costs that are upfront, and that can be a lot of expense long before revenue is coming in. Build vs Buy is always a difficult decisions for companies, but being focused on impact on BOM and ignoring all of the other costs, both in time and money, is dangerous.
The last thing I think a pinball company should do is create a new board for each machine, in order to have just the right amout of 'outputs' . Look at the frustration with Stern Node boards not being compatible from machine to machine. With P-roc, for example, you can take any board from a TNA and use it in a Houdini or P3, and vice versa. Obviously boards will evolve, just like the heck board has, and while it will not always be compatible across all machines, currently it is across a few different models. The only unique boards you would want to engineer are those that are very specific to a game. Even on Houdini, we built two custom boards for electronics, but both were built it so he could handle more 'stuff' then we needed, so we could re-use them on future games.