I have no idea how to solve your specific problem but I do know that you can fake it if you’re really desperate; you just won’t be able to edit part of the output in MIDI.
You can plug the “Phones” our “Line-Out” jack of your synth/keyboard/etc. into a USB audio interface and also plug the Midi/USB cable into your computer. From there you can just do a simultaneous recording session where you record one track as a MIDI track using the VST on your computer and the other track as a WAV Audio input from your USB Audio Interface.
That will give you 2 full recordings: 1 with the WAV Audio information directly from your keyboard/synth and one with the same thing as MIDI information, obviously using the VST from your computer.
From there you can just cut them together. It would work fine if you didn’t plan on needing to do very much in MIDI after the fact. The main downside is that you’ll only be able to hear one or both of the voices during recording which could make things difficult or even impossible depending on what you’re trying to do.
You could always just do two separate recording sessions and stitch them together. I wish I had a better answer but something’s better than nothing I guess.