Thank you all for your ideas and help. I will try to explain in details, so that you can find answers on your separate questions in it, ok.
I wrote a story, script (along with the dialog of course).  There are 4 talking characters (actors) in the movie. It is short animated movie (around 30-35 min) in non-english language with English subs. Scene are done in Unity HD project, those without dialog (already finished and rendered in Unity in 4k scope (2.39:1) 4096×1716 in png image sequences) are in post-production stage for video and sound editing in DaVinci Resolve. Most of the sfx, folley and music is already done. Characters were animated in Iclone (using CC3 female base, CC2 male base, Hivewire Dusk and DAZ Genesis 3 male , and lot of animals models). I had a problems with fbx animation export to Unity at first, but when I start using alembic these problems were eliminated, with more time consuption, clearly. And then time came to make these dialog scenes, which I left for the end.
I didnt want to start recording voice of the real actors (real people who showed interest in cowork with me on this project) separately, because I was still looking for the proper workflow in which I can record dialog (voice) and face expressions at the same time.
Of course that I could record voices separately and then (even through Iclone alone) tweak it and lipsynch it somehow, but I am affraid that this would ruin already smashing look of the work done so fare (you can never get convincing face expression/talk flowless look).

So I first tried with Cinema Face Cap - Facial Capture for Unity. For that reason I got this Kinect v2 for Xbox One with Microsoft adapter (Kinect is working great by the way,in Win 10 with SDK from Microsoft, and Brekel ProFace 2 (trial) with 4 seconds of recording time for free (before I purchase) but already had problems  with Iclone models in Cinema Cap (guy from customer service asked me for one model to upload them for testing, because they didnt have experience with them sofare, still no answer).  The way it works is you capture your face talking in realtime inside unity and record that animation, which sounds great).

In the meantime I tried to apply facial data through DAZ Genesis 3 male. The animation was looking good, and at that point I was thinking  that I if I could export that to Iclone and convert to Iclone model, then I could maybe copy every new animation to other characters from Iclone (after succesfull  conversion from 3dxchange). Thats where I stucked.  It could be done with Motion Builder I know, but I dont have time to learn MB now, so I would like to stick to Iclone, which is familiar to me.
When I export it from 3dxchange as iProp, everything is fine, so at least I can make close-ups with just face in it (if you know what I mean). This way everything works, but then loosing iAvatar animation tools, clearly.
Thanks for your help and ideas. I will find the way, I am sure, I was just hoping that 3dxchange will fix this issue, but it seems it will never do that, so I guess, to continue my workflow inside the Iclone with fabulous alembic export to Unity, I will eventually have to go with Live Face or Faceware.