An experiment with phonetic writing

[eggyk]

Isn't the premise of this that the french person would potentially understand the word once re-spoken by the person transcribing?

It’s strange that you’re stopping at this after I’ve confirmed it. Yes, a Frenchman will understand this person if he says “aujourd’hui,” and he will understand this person if he reads it in his letter. But he might not understand what he wrote.
By analogy with Russian: There is the word “Солнце”. It is pronounced almost like “Сонце”, but some people might hear the sound “a” in the last unstressed letter and write it as “Сонца”, or even “Сонц”.
How quickly can I recognize that Сонц = Солнце? Well, maybe quickly, because if you write Russian with some mistakes, it will still be understandable. As far as I know, French can be made completely unreadable with these mistakes.

I've only just joined the conversation 

What is the point that you are trying to make here? That someone writing phonetically may produce text that others cannot read? Of course that is true. 

But this all depends on how accurately the transcriber interprets the sounds, exactly how his system works, and exactly how much the other person understands the system (or how good they are at pattern recognition). In order to reasonably try to write phonetically, the transcriber must surely be reasonably able to interpret the sounds of the language, at least to a point where they would feel able to do so. 

So, if the purpose was to phonetically write down french words in order for a french person to read, why would someone do that? Why not just learn french spelling, and then write the book? I feel that this breakdown in the metaphor is a consequence of using a language like french, instead of a completely different writing system like chinese.