Although "everyone" knows what you mean when you say "voice" - Yamaha spent a reasonable amount of engineering and marketing effort to get rid of "voice" mode in order to provide a more simplified/elegant solution. Therefore, you will be met with some "terminology correction" feedback early on.

Now getting past that - since even using "wrong" terms - we can all arrive at the same place. Take a look at:

Source: Yamaha
https://www.youtube.com/watch?v=JXASiv3YfSE

Source: 3rd Party
http://www.moessieurs.com/user-waveform_montage.html
(page has button to translate from French to your native language)

Also, a Montage-specific version of the MWE waveform editor (3rd party) is in the works. This would allow for offline assembly of a Yamaha-specific "waveform" which would be a multi-sample entity with several parameters embedded. In the interim, you could purchase the Motif XF version of this utility and use it to generate Motif XF files which can be read/converted by Montage.

There is no specific information that I could find on the expected release date or upgrade path if you get the XF version and a Montage version is later released.

Also, for the 3rd-party offline editor route - Motif XF also had more than one 3rd party offline waveform editor (more than just MWE). There's no information on "the other" solution with regard to Montage in terms of eventual support for Montage. However, presumably you could also use this tool to create Motif XF files that can be imported to Montage.