Just to put it in simple terms what did the author accomplish by adding the configuration to a Vyatta VM?
Warm fuzzies?
In this case it is not about nested VMs and from your reply this option only restricts the usage of the backdoor, so what is restricted as I am not familiar with "guest IOPL >= its CPL"?
CPL is "current privilege level." OS kernel code typically runs at CPL 0 (the highest privilege level). User application code typically runs at CPL 3 (the lowest privilege level).
IOPL is "I/O privilege level," and controls the privilege levels at which I/O operations are allowed. If the IOPL >= the CPL, then I/O operations are allowed. Otherwise, they are not. Under most operating systems, the IOPL is almost always 0, which restricts I/O operations to CPL 0 code (i.e. the kernel). With the monitor_control.restrict_backdoor option, hypercalls are similarly restricted.
So, in typical usage, this option restricts the use of the backdoor to the guest kernel. However, depending on the OS, it may be possible to modify the IOPL of a user level process. Linux, for example, provides the iopl() system call, which can change the IOPL of the calling process. Any process which can raise its IOPL to 3 will be able to make hypercalls, regardless of the setting of the monitor_control.restrict_backdoor option.