From my experience with non-automotive networks, the other big issue is the availability of components that can communicate using a particular protocol.
For example, the very expensive and highly sophisticated grid-tied inverter I purchased just 4 years ago as part of my solar PV system, can only communicate with the outside world using low speed serial RS232. So in order to collect production data from my inverter, I have to use a RS232 to USB adaptor and a dedicated netbook sitting in the garage. By contrast, my Blink charger is on my home WiFi network and updates the data automatically without me having to do anything other than use it.
My point is that if all the available components only talk CAN (or USB 2.0 or SCSI or RS-232) the car maker is stuck with that protocol until component makers start making engine controllers, instrument clusters, ABS sensors, etc. that can support the newer protocol. MB can afford to design make all of its own stuff or force Bosch to make MOST components so they don't have to worry about component compatibility, but Fisker is not quite there yet.
I agree in concept, but in this particular case I don't think it applies. MOST is not a proprietary MB mechanism. MOST technology is used in almost every car brand worldwide, including Audi, BMW, Hyundai, Jaguar, Land Rover, Mercedes-Benz, Porsche, Toyota, Volkswagen and Volvo. (Lifted from the Wikipedia article)
It's not all that new either, even the newest iteration of MOST, MOST 150, is almost 5 years old. There's no shortage of suppliers and maufacturers that make MOST-compliant products.
Of course, when I killed the audio system in my Mercedes (eventually traced to a shorted head unit) it took a village to troubleshoot it because there were several suspects on the MOST ring, plus the MOST ring itself to consider.