This is a joke, not usable at all.
Their new dictation stuff is quite good, just not for your specific application (Medical Dictation). It does work well fairly well for general dictation.
There is no way to implement a special, in our case medical, vocabulary.
There is a way to do it, it is jut not easy for an end user.
This "embeded" crap is, may be, good for a housemaid (nothing against housemaids), but not if you have to work with it as a professional
Have you tested it recently for general dictation? Not your speciality use? If not, how are you making generalizations about its usefulness for others.
Belief me, we use OsiriX MD (an OSX application) to review medical images and Dragon on a Parallels VM to dictat the reports in day to day work.
I am not surprised that it does not handle dictation for the specialized market of Radiology. I am curious if you have tried VoiceboxMD? It is dedicated medical dictation App that runs on Windows, macOS and iOS and seems to be about half the price of Nuance’s offering. Radiology is one of the specialties they specify.
Also, do you know if Nuance’s Medical One will run on Windows on ARM? It seems from their specs that it can run on 32-bit Windows systems, that might mean that it would run on Windows on ARM (that might run virtualized on a Mac).
We will replace OsiriX MD with a Windows app when the next of our Macs is going to hell.
As we did it with several of our Macs / the XServe in the past.
I would hope you would look at the options that are available for you at the time, rather than making a decision now with incomplete information. However, it is certainly possible that your needs will no longer be met by the platform in the future. It is also possible that support for the new Neural Engine might make it much easier for someone to do a medical dictation app that runs native (and does not require a cloud connection). Unless your systems are about to die, why worry about it now?