Is iota ever used as a consonant as it is in Latin, or is it always considered a vowel?
It’s a vowel.
When it appears as a consonant in the poets — and it’s not common even there — it’s almost always to bring a metrically difficult name into the verse.
I am not too familiar with the linguistically correct way to pronounce ancient Greek of an earlier period than the Koine (I know most of the theory but haven’t really practised it and there are gaps in my knowledge base including the pronunciation of iota 
)
In Koine however iota can sound like a semi-vowel and that is why it has been used for the transcription of such names as Yosef (just put the right accent marks on ‘o’ and ‘e’).
It would be interesting to know how Christian poets using the old meters handled names starting with iota. In the classical language a name like Ἰάσων “Jason” has three syllables.
annis I agree it is intriguing (as far as I know iota always sounded like a vowel before; I just don’t know for sure, ergo my the beginning of my previous post) .
While I am not an expert on the matter, I have to tell you that it is even more difficult for me to understand why poets in the Byzantine Empire (when the pronounciation of the Greek language was almost exactly as the modern Greek one -ypsilon being the most notable exception-) still insisted, for the longest of time, to use the metric system of ancient Greeks. It just doesn’t work!
So what is a consonantal Iota?
English y.
Your short and to the point answers usually work but this time it passed me by .
It is pronounced as the English y but how was it written in Greek when it is pronounced that way.
Looking at Smyth 20 I see that an initial consonantal Iota sometimes passed into rough breathing or Zeta. Medials were often lost.
Are there any examples of an Consonantal Iotas in Greek words without these changes or was that from an era of which we don’t have any records?
In the historical period it would not be distinguishable from a normal iota.
Looking at Smyth 20 I see that an initial consonantal Iota sometimes passed into rough breathing or Zeta. Medials were often lost.
These happened during the transition from proto-Greek to the various forms we see in the historical period.
Are there any examples of an Consonantal Iotas in Greek words without these changes or was that from an era of which we don’t have any records?
Linear B seems to preserve consonental iota (at least to my eyes, that may be a artifact of the writing system), but y was gone by the time Greeks adopted their alphabet. Actually, some of those initial consonantal iotas were already zeta in Linear B.
Is there any reason why the iota became a zeta?
If I remember what I’ve read, this is one of the problems of Greek linguistics, why some consonantal iotas became zeta and the others became rough breathings. Wikipedia has an article on Proto-Greek that mentions it (although maybe I shouldn’t reference that particular source ), that seems to offer a solution to the problem.
But the same thing (or something very similar) happened in French (and partway in Italian, but I’m not sure about other Romance languages), where you now have jeter, janvier, etc. with the zh sound.
Personally, I find that it’s sort of easy to turn y into z by repeating it over and over again. If you do that enough, the y starts to take on characteristics of a “dz” or “zd” sound.
So was
)Ihsou=j
pronounced with a consonantal iota?
Well it’s a bit more complicated than that. Since quite often the iota at the beginning (and in other cases but let’s stick to the beginning) was/is pronounced as the english ‘y’, it was used even in cases where this was not the case.
In the case of Jesus’ name, it is not pronounced exactly like ‘y’ nor as ‘i’ in ‘pit’. Think of a “soft” ‘y’ (something sounding closer to I than y).