There's 2 problems with this.
1) Business customers will not get any comfort from "this might not be the best way to do it". What they want is certainty that they can spend money on pro-level mac hardware for their software development teams and it's not a wasted investment. Until there is certainty them - for sure - those purchases will be on hold.
2) There's this assumption that because OS X is bigger now, Oracle will jump in and support it directly. Ain't neccesarily so. Consider: Java is mostly prevalent in the server space - and Oracle has a big linux investment, so JVM on linux makes sense. By covering that and Windows, they give a huge set of platform coverage for the development side. Now, consider - if Oracle don't support OS X - how much money are they going to loose? I'd say 0. It is *far* more likely that enterprises will change their hardware platform away from the Mac than they suddenly stop writing Java. Why would oracle spend time and money on a port when all they need to do is wait; Java users have to have a supported development environment so they're forced back to Windows / Linux.
Java application haters just don't get it. A huge amount of software development targets Java, only a tiny fraction of it is used by consumers. Those guys need computers, they liked OS X because it's got good java support, UNIX, and a nice GUI. There's been a huge uptick in macbook toting IT people, who then were a great ready audience for writing iPhone apps because they had the kit. They don't care what the app store rules are, they don't even care if Java is an OS component or a separate download; or even who supplies it - just the assurance that it is there and supported. An no, some shonky Soylatte build on X11 isn't going to cut it.
Also, you may be surprised what consumer stuff IS java. Flash builder; eclipse IDE. Moneydance personal finance; Java. PS3 media server; Java. Many language development environments are eclipse, and therefore java, based.
No JVM = No OS X = No enterprise apple sales.