An experiment with phonetic writing

[ololololo]

If they are pronouncing it the same way and writing down the same pronunciation in different letters, what is the inaccuracy here:

I’ll provide the same text, written down from hearing it:
Translate:
Bonju! Sava?
Sava biean, etua?
Sava biean. T'qua tcokuptu?
Jevuea epari sesuar, etua?
Jevuea mepromene.

In terms of how it sounds, “je” and “vais” merged into “jevuea” (well, if a person is smart, they can separate “je” and “vuea”). It doesn’t look much like French, does it?

This doesn't seem to me like you are saying the accurate pronunciation is mostly getting recorded,

Yes, the correct pronunciation is recorded. But it’s not guaranteed that it will be conveyed correctly (ujuvi, oujovey), and it’s not guaranteed that such a text will be understood by a native speaker.
This is if we’re talking about writing in Latin script, i.e., when the listener will be able to read our scribbles. According to the theory, the author of VMS acted differently.

I am just trying to understand who this person who can hold a conversation and design an alphabet for distinctions in phonology but not hear them it in dictation is supposed to be.

Well, here’s how I see the situation: the author created an alphabet for sounds and could clearly distinguish them in words (I think there was still some accuracy, at least in terms of tones). However, he didn’t know the language perfectly; he studied it on his own, and even then only through conversations. He could roughly understand what was being said to him, but that didn’t mean he would correctly understand all the sounds or all the words.
Here, it’s easier to use an analogy, for example, with English in Russian schools. If you don’t study it additionally, your school‑level knowledge will allow you, for example, to read and somehow understand a text, but you won’t be able to have a normal conversation with a native speaker (because you’ll be a very boring conversationalist, slowly forming phrases). In our case, it’s the opposite — you’ll be able to speak, but you won’t be able to read a text. 

I hope I explained it clearly.