I think a little historical perspective may be helpful here?
There have been many cases where the "in house" driver was "the thing". The home-grown technology was the differentiator. Pioneers of electrostatic, ribbons, AMTs, etc. placed them on the radar -- and, kept them there. A Japanese company that builds motorcycles and pretty decent musical instruments developed the berryllium dome in house. This is just one example.
Then there’s the case of the Dahquist DQ10. An off the shelf woofer used successfully in a mid-range bookshelf speaker along with an array of "nothing special" components -- including a $5 Motorola plastic piezo super-tweeter. It was the attention to "other" factors that made the speaker a future Audio Hall of Fame’r. More recently, Golden Ear has had success bringing vintage technology that has been somewhat domant back into the spotlight. And uses off the shelf drivers from "others."
I can think of may factors why amanufacture would want to make components "in house", possible supply chain issues being one of them. To me, it’s not the cost of components that is the major factor. It’s whether or not the end user gets their money’s worth at the end. A measure of commerical success is usually a prettty good indicator.