Already discussed and in practice among those on FB. Thought it was already discussed here as well.
In a nutshell, the cost of running the plasma for remote based operation is much higher than just using a remote. While the cost in electricity may not be much of a concern, the wear and tear on the plasma tube may be a factor to some.
The remote can do repeats without blinking an eye. Plasma would be a waste to run in repetitive mode.
Where use of the plasma's scalar output for remote does become useful is when the output of the program is plasma specific. The SC's modulation mechanism can not be 100% replicated by the generator by itself. So if you have a particular plasma specific frequency spread due to sideband creation, then it gains some merit.
However, being in the far field would probably still be better than relying on remote transmission. I rather pause the plasma while running to the store than rely on the remote aspect. Just my preference.
However, adding a remote component while also using far field may be worth investigating.
One other area where it comes in handy is if you are trying to apply plasma specific outputs to targets that can not get in range of the plasma. If all you have left is a remote application, then by all means. I usually only add remote targets to plasma sessions though when there is actually a local target for the plasma output, never just for the remote target.
Finally, it does work and works well. I have invoked responses in remote targets via this method that mirrored responses invoked via normal remote methods.
For more details, please check the link: