Voynichese is a numeric cipher?

[ololololo]

Second, let's look at dchdy - two letters d. This is very similar to terra. Therefore, d = r, which means that ch and y should be equal to e and a. It turns out to be a literal translation.

dchdy with d=r, ch=e, y=a is:    "rera", not "terra"

i-g-e-r-e - it sounds like the word "inrigat" (to water), but we don't have t, a, n and the second i. Is it a shortened version of the word inrigo?

"igere" does not sound like "inrigat" or "inrigo" in any meaningful sense. "igere" also sounds like "belligerence" without the b, an e, and the l's. 

is-n-e/a-s - this is reminiscent of the word ventis (winds), but we are missing one letter.

"isnes" or "isnas" is not reminiscent of "ventis". Even with the v it's a really big stretch.

Has this thread become an AI driven theory thread?

Oh, well, thanks for the comments! Ch = te, not e. 
As you correctly pointed out, this interpretation has assumptions and is not entirely accurate (as I mentioned in a subsequent post). And yes, I am also mistaking 
But I don't pay attention to it...
And yes, you've noticed that the words don't look like irrigat or ventis. That's the only thing I've found that matches. But this isn't a translation, it's just something I've picked up and tried to make sense of (the translation should be aqua and aer, logically).
I am absolutely certain only about otol = ignis and okalo = intesti.
I'll also explain about rera. So, yes, dchdy = rtera (ch = te), but by rearranging one r, we get terra (if ch = e/t, then it becomes trra or erra, which is not too far from terra).