Monophthongs in IPA

Hello all,

The pronunciation of the three pairs of monophthongs is
described by Warder thus:

a  like u in 'hut'
aa like a in 'barn'
i  like i in 'bit'
ii like ee in 'beet'
u  like u in 'put' or oo in 'foot'
uu like u in 'brute'

In IPA these would then be:

a-va.n.na:     [inverted v], [script a:]
i-va.n.na:     [I], [i:]
u-va.n.na:     [upsilon], [u:]

I am curious to know what is the basis for the judgment that
the two items in each pair are phonetically distinct (as
opposed to being two identical sounds that differ only in
duration) ? I don't see any evidence for their phonetic
distinctness in the Pali grammars that I have looked at; so
is this a conclusion that has been arrived at by comparison
with Sanskrit, or on the basis of modern pronunciation, or
something else ?

Is there any possibility that in Buddhaghosa's day the
monophthongal system may have been like that of Icelandic,
in which every vowel has both a long and a short form,
according to which and how many consonants follow it ? I am
thinking of something like this:

a-va.n.na:     [inverted v], [inverted v:]
i-va.n.na:     [I], [I:]
u-va.n.na:     [upsilon], [upsilon:]

or:

a-va.n.na:     [script a], [script a:]
i-va.n.na:     [i], [i:]
u-va.n.na:     [u], [u:]

Or perhaps some blend of the two. If this is implausible,
then what exactly is the negative evidence that counts
against it ?

One other question: from the account in the grammars it
would appear there are several possible contenders for the
vowel 'a' besides [inverted v]. I'm thinking in particular
of [turned a], [script a], or even [schwa]. What is the
basis for the preference for [inverted v] ?

Best wishes,
Dhammanando