1. Alif MaqSoorah now and in early-Islam times
As it is well known to all who have some knowledge of Arabic, the orthography of
Classical Arabic recognizes the rendering of the long wowel "a" (or as I often
may transcribe "aa") by the letter Yaa' (in texts with vowel signs and
diacritical points: after the vowel sign FatHa for "a") -- but only at the end
of a word, like in dhikraa, kubraa etc. This is termed an Alif MaqSoorah.
Traditional Qur'an orthography (as represented by the Cairo Qur'an), however,
retains the Yaa' even if the "aa" is not in a final position, but is shifted
into a "medial" position by the addition of a suffix to the word, e.g. in
ra'aahu. The usual explanation of this phenomenon is that it is still to be
considered as an Alif MaqSoorah for etymological reasons, since these words are
derived from verbal forms like ra'aa e.g. with the vowel "aa" at their end.
Actually now there are a number of old Qur'an manuscripts under scrutiny the
oldest of which possibly go back to the early 8th century or even the end of the
7th century, though it is uncertain whether they really will enable us to regain
a text form of the Qur'an earlier than Uthman's Qur'an redaction. I may term
them "Hijaazi manuscripts", because they are written on parchment in the Hijaazi
style of script. They are (of course) without diacritics, not vowelized etc and
often show "defective" writing of long vowels. But they also show the
Alif-MaqSoorah writing of "aa" in many more instances than would be correct
according to the traditional Qur'anic orthography or even the orthography of
Classical Arabic.
2. Writing of "ilaah" with Alif MaqSoorah
Especially, these Hijaazi manuscripts occasionally write the word ilaah "god"
with the rasm (i.e. approximately: consonantal script) "'lyh" or
"Alif-Laam-Yaa'-Haa'" instead of simply "Alif-Laam-Haa'", as in the Cairo Qur'an
and as it is usual. The letter Yaa', no doubt, is used as an Alif MaqSoorah, but
without any etymological justification like in the cases mentioned above.
For the experts I add: We have good reasons to reject the idea that this Hijaazi
orthography represents the variety of an Imaalah-pronounciation of the
"classical" sound of "aa". Thus we are to deduce that the Alif MaqSoorah for
writing the long vowel "aa" was one of three or four options the Arabs had at an
early time, whether there were etymological justifications or not: the oldest
being the "defective" writing without any sign for the long vowel, the youngest
and succeeding one being the "plene" writing with Alif, partially in use till
today is the writing with Waw and and intermediate one being the "plene" writing
with Alif MaqSoorah, i.e. Yaa'.
3. Ambiguity between "ilaah" and "aalaah"
The spelling (rasm) "'lyh" was even involved in finding the ultimate orthography
for God's name "Allaah" in Arabic. It is hard to believe that "Allaah" goes back
to Arabic "al-'ilaah", as Edward William Lane summarizes the traditional
theoretical efforts of the Arab grammarians in his Arabic-English Lexicon. The
derivation from Syriac "alah" is much easier from the phonological point of view
(For the discussion of this aspect see Arne A. Ambros "Zur Entstehung der
Emphase in Allah", WZKM 73 (1981), p. 23-32). It remains to realize the problems
the Arabs had in those times to express the name "aalaah" in Arabic script.
The genuine Arabic word for the generic denotation of (any) god is "ilaah". A
close transcription of Syriac "aalaah" into Arabic therefore resulted in
exactly the same spelling as "ilaah". The only differentiation could then be
achieved by the pronounciation or by analyzing the context. Both would allow for
a longer period of an identical expression in script of "ilaah" and "aalaah".
Actually, the early Hijaazi Qur'ans due to their older orthography hardly
contain any verse without an ambiguity in script. The elimination of blunt
ambiguities like "ilaah"/"aalaah" certainly was an urgent part of a general
orthographic reform which, in this case, resulted in the later usual, a bit
complicated writing for "Allah".
http://answering-islam.org.uk/Quran/Sources/alaha2.html