The Openness of Open systems

I was having a spirited conversation with a friend (involving beer of course) yesterday evening. We got around to discussing the sham of "Open Systems" - especially as applied to the travel industry. My belief is that the Open Systems marketing term was coined to attempt to break the stranglehold of TPF (transaction Processing Facility) on IBM mainframes. Now TPF is an unbelievably reliable system, but without a lot of flexibility. Maybe that's what makes it reliable. So when you want the system to be rock solid, run on TPF, otherwise do something else. That all comes at a price, so "Open Systems" were used as a way to beat up the TPF crowd with the promise of cheaper and more agile computing - and supposedly the promies of being able to switch - i.e. not be beholden to any single vendor. So *NIX systems were the darlings. Well kind of. Solaris, HP-UX, AIX were all candidates. But realistically how easy is it to switch from one flavor of OS to another?
That led me to draw an analogy between Operating Systems and DNA. A pretty woolly analogy, I admit, but it came out kinda fun. Let's imagine there is somehow a bit less than 2% difference (whatever that means) between a pair if UNIXes (UNICes?). That's about as different as the DNA differences between humans and chimpanzees.
It doesn't take a lot to see that the ~2% difference at the DNA level leads to some pretty major incompatibilities. That's true in the "Open Systems" world too. 2% different? Not much will switch easily.
So I continue to suggest that the openness of "open systems" is still a crock.