Knowing mainframes though (at least many years ago) I can tell you that IBM has been virtualizing its hardware for its mainframe architecture for so long that it wouldn't surprise me a bit that their patent portfolio practically locks competitors out of emulating it.   Their tech for virtualization goes back to the late '60's and early '70's and the 360 line of computers.

VM, (Virtual Machine) was one of their first mature operating systems.   Their LPAR technology (running multiple operating systems on one set of hardware] is based upon that tech.   In the early 90's they produced  "mainframe on a card" for developers which emulated the mainframe in some custom 68000 chips.   In the 90's Fundamental Software did the emulate thing for Unix and IBM hit them the same way, resolved by IBM basically buying out Fundamental.  Some of their newer low end mainframes are actually POWER architecture running emulation (don't know whether they used any of Fundamental's tech or not).

So IBM has a long history of protecting its mainframe turf.  

The only counter example is the open source Hercules project which emulates mainframes in software.  But last time I looked the license made it very plain that the project could not be used for commercial applications.

The "evil" here is not with the companies but with the patent system.  If patents protected only hardware, IBM wouldn't have a leg to stand upon.  But as we do allow patents for software, then I suspect it is indeed impossible to emulate the recent incarnations of the mainframe without running afoul of IBM's patents... ergo the patents are doing what they were intended, allowing an inventor/innovator to protect their government granted monopoly.

Back when I was a mainframer, I bitched bitterly that we needed the ability to run mainframe emulator's and current software at home if there was to be a future for the mainframe operating systems.  The XT-370 ran about 20K with very limited licenses.  I was "wrong" only in that the z/OS operating system is going to run a long time on sheer inertia, and VM was "saved" by Linux ... but many features of the VM stack (CMS, REXX, and Pipelines in particular) are quickly withering; it is only the core hypervisor technology of VM which is getting any attention and has any real future.

As I moved to Unix my perception changed.  While I loved the VM environment and saw a lot of strength in it, I realized that basically time has passed it by.   IBM's perception may have been better than mine was.   On the other hand if we stubborn petulant VM'rs hadn't been so vocal and made such a nuisance of ourselves and kept the VM operating system stack alive despite many attempts within IBM to kill it, it might not have been there and still viable when Linux came along. It's quite possible that the early amateur discussion and implementations of Linux on VM (in which I played a small, purely jawboning, part) had some influence and impetus on the "skunkworks" project which eventually became the official Linux on mainframe release.