Some VR platforms allow you to display the contents of a monitor, or even a specific window, in VR. If that functionality is not implemented natively for your platform, there are also "virtual desktop" applications that do the same thing.
Individual words can be added to the speech recognition dictionary (there is a shortcut to that feature in the context menu for the "Utilities >" button on the "Recognition" tab of the VoiceAttack options window), however that is not a replacement for a properly trained speech recognition profile.
As mentioned, short words are the most difficult for the speech recognition system to recognize accurately, as there is much less information to work with.
It's also worth pointing out that the speech recognition engines are designed for a specific accent. E.G. if you don't speak English with a US accent, the US English speech recognition engine will likely not work accurately.
If there is a speech recognition engine available in your native language, that may be worth trying.