Allah is الله ʔallāh, and it does start with an alif. It's also true that Modern Greek renders it as Αλλάχ. That's just about all the OP got right.
First of all, the comparison with Hebrew אֱלֹהַּ ʾĕlōah (actually an analogical creation based on the irregular plural אֱלֹהִים ʾĕlōhîm of אֵל ʾēl) is nonsensical: yes, it starts with an aleph, just as Allah starts with alif, but what the OP doesn't seem to realise is that aleph and alif are consonants—glottal stops in both languages—and what is being rendered in Greek is actually the following vowel. The glottal stop was never phonemic in Greek at any stage and at the start of a word wouldn't even have registered as a distinct sound to a Greek speaker (as indeed it clearly didn't to the OP), so it isn't reflected in writing.
(As an aside, Mark 15:34 actually has ἐλωῒ ἐλωῒ (λεμὰ σαβαχθάνι), not ἐλοΐ ἐλοΐ, and Matthew 27:46 has ἠλὶ ἠλὶ (λεμὰ σαβαχθάνι). Both are probably Aramaic (the rest of the quote certainly is), not Hebrew, but the Aramaic also starts with an aleph. Psalm 22 starts with the Hebrew equivalent: אֵלִי אֵלִי לָמָה עֲזַבְתָּנִי ʾēlî ʾēlî lāmā ʿăzaḇtānî.)
There is no reason to expect initial ε- in a Greek adaptation of Arabic ʔa- at any stage. In Classical Arabic a was generally realised as [a], just the same as Greek α at every stage of the language, and we would expect it to be rendered as such. We obviously don't have a ton of early Arabic loans into Greek, but Herodotus, in the 5th century BCE, actually already renders the name of Allat, a pre-Islamic Arabian goddess, as Ἀλιλάτ (presumably reflecting *ʔal-ʔilāt, which later became ʔallāt, just as ʔal-ʔilāh became ʔallāh).
The only place where Arabic ʔa- might seem to turn into ε- is in Modern Greek εμίρης 'emir', from Arabic أمير ʔamīr, but that's mediated by Ottoman Turkish emir; in Byzantine Greek (that is, the Greek of the Middle Ages), we actually have a direct loan from Arabic in ἀμηρᾶς /amiras/, also with α.
Engaging with the argument on its own level, there's also the point that some of our earliest manuscripts of the relevant passage of John (Papyrus 115, the earliest one, and the Codex Ephraimi Rescriptus, among others) actually give "the number of the beast" as 616, not 666. The Church father Irenaeus did affirm 666 as being the number of the beast in the 2nd century, but if we're taking Irenaeus as authoritative, we should also note that he strongly condemned exactly the sort of numerology the OP is engaging in as an absurdity in his Against Heresies. (He goes into some detail about how reducing Jesus to the number 888, as in the OP's username, is blasphemy, in particular.)
On a more sensible level, the point could be made that obviously John wouldn't have known about Allah either as a word or as a god, but taking a stance against prophecy being real is probably a non-starter in this sort of discussion, I guess.
None of it is worth devoting any attention to, to be honest. The OP is trying to rationalise his preëxisting Islamophobia by throwing things at the wall, not making a logical argument.
