Oracle is successful, due to a collection of great technology but with Operating Systems they have obviously not got some facts still
(and hence they have less maket share than they could have):
1. people do not use Oracle server hardware as desktop workstations.
2. 99% of all Solaris users are not responsible for decisions on buying hardware
3. The eco-system is the primary driver to establish market share, not marketing
4. Feedback is important and must be enabled (and in beta programs must be free of charge)
All kinds of interested people will try Solaris 11.2 beta.
A bad impression will be, if it fails while - or right after - booting.
Unfortunately most people are forced to use some kind of x86 hardware nowadays.
Unfortunately again, contemporary x86 hardware is not compatible, since it uses some kind of GPU, either AMD chipsets or Intel Haswell chipsets
which both need a kernel driver.
Although Solaris is a basically a great server system, it has always also been also a great OS for workstations,
like in CAD or other fields of engineering.
I think Oracle has enough skilled engineers to grant Solaris some more focus on X, on future architectures like Wayland and on resp. device drivers.
This must be done pro-actively together with the chipset producers
(instead of waiting until a Linux driver comes up and can be rewritten to be usable in Solaris x86).