About your French VCV inquiry – it has been done before. There is actually a French VCV reclist here on Utaforum, created by Melomad:
https://utaforum.net/resources/gros-mot-french-vcvvc.375/
Beware though that as far as I remember the reclist contains not CCV samples and instead would put them together out of simple CC recordings (like recording just “kr” without a vowel), which in my opinion is sup-optimal. They should at least be recorded as “kra” or something else with a vowel to sound more natural and work properly, similar to how Cz does CCV clusters in her VCCV reclist.
A friend of mine wrote a private extension list which adds CCV to the list, maybe you could contact here if you want to know more:
1. What languages have I seen VCV reclists for?
Many. Of course there are Chinese, Spanish, Korean, even English (masochists, all of you!), French, Finnish, and even German and Vietnamese, and probably more that I just can’t think of from the top of my head. I wouldn’t be surprised if there was a Russian VCV somewhere in that part of the fandom.
I’ve started writing a VCV+CCV+VCC reclist for Latin once, also one for Serbian. A long time ago.
2. Is it understandable?
They usually tend to be understandable, but it also depends whether they are purely VCV or if they have CCV and VCC extensions.
3. What language would I want in VCV?
As others have stated before me, VCV is best suitable for a language that knows few consonant clusters, be it CCV or VCC. Even better if the language only have a couple of vowels and not too many consonants, and another added bonus if in a word consisting of multiple syllables, there are little in-word-clusters and more often than not CV-CV type of combinations. Which means that languages like Japanese and Hawaiian are best suitable for a VCV type reclist – few vowels, few consonants, little to no consonant clusters. It keeps the reclist comparatively short. Spanish for example only has 5 vowels, but a decent amount of consonants and there are also clusters – these clusters are handled in different ways by different reclists, so some record the clusters, others leave it up for the USTing.
Some languages can be an awful load of work even in CVVC format, depending on how thorough you are. English and German are – from my experience and with my way of writing a reclist – are *hell*.
Going the VCV route with them… one really has to think about what they’ll include and what they’ll leave up to UST editing. E.g. not recording diphthongs in the same way as monophthongs and instead putting them together via vowel blends in the UST. Or not recording CCV and/or VCC, instead leaving that up for UST editing.
Even then, although writing a VCV reclist for almost any language is certainly possible, and there are of course some people who are willing and able to record and configure all of it, a VCV type reclist isn’t always ideal for a typical UTAU user.
The recording strings can be too long for one’s breath, and shorter strings will naturally result in more lines to record. For that, many do not have the time, energy or patience (others are *living* for that). The otoing workload is immense as well, taking up a lot of time one may not have and could have possible instead recorded two CVVC pitches instead. Depending on the language and reclist, more than one pitch might not be feasible as the work and voicebank size would be just too much.
I admire those who stick to VCV for pretty much any language and are just doing it, having the energy and patience for it. You all rock! (…I still think you’re masochists to some extent). And I also admit that writing a VCV reclist can be fun, at least the theorizing part.
I nonetheless think that VCV works best with a few languages of a certain type and that other languages work better as CVVC on which one can add various extensions.