But I'm no journalist... and Nathan I think you nailed it! As you often do.
Now who comes to a Linux news site to talk about the virtues of QNX. I don't see your line of reasoning honestly. Android does not need an easier route into automotive and QNX does not seem necessary to achieve it even if it was needed. (As pointed out, Android is already in automotive in a multiple of different configurations). If isolation and containerization is what you desire, you can do that with Linux. If you want to take advantage of the quick advances in a modern Linux stack (systemd, Wayland, kdbus and friends, and the automotive standards coming out of GENIVI) then this is challenging enough as you may still be ripping out large parts of the Android stack even when forking Android while staying with the Linux kernel. But how would taking on such a task on a different operating system with a completely different kernel possibly make it any easier?
IMHO, the only reasonable long term approach is a fork of Android "on Linux" if a fork would be at all necessary, but in reality that might not even be needed if device connectivity picks up the slack instead, as the OAA site and this article alludes to.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds