>Phonology - Consonants:
>          Labial     Alveo-Dental     Velar     Palatal     Glottal
>Stops      p  b          t  d         k  g                     ?*
>Fricatives   v            s                      c /S/         h*
>Nasals       m            n             ng*
>Flaps                     r
>Laterals                  l
>* ? and h only occur initially; ng only exists when pronouncing -nk/g-
>Phonology - Vowels:
>          Front     Central     Back
>High        i                     u
>Middle      e                     o
>Low                    a
>Diphthongs: ai, au, eu, ia, iu, io, oi, ui

Well, this far you've already explained us that "?" and
"ng" are not real phonemes in your system, so I'll start
my comment from your actual consonant chart:

Stops: p/b, t/d, k/g
Fricatives: v, s, c, h
Nasals: m, n
Flaps: r
Laterals: l

First, I find that contrasting flap vs. lateral is not a
very good idea, since their acoustic difference is small
and to e.g. a Chinese ear a flap sounds more similar to
an l than to a trill. If you really want your chart to
contrast rhotics from laterals, my advice is that you define
/r/ as some kind of trill --you have a handful of them
to choose from: the lax retroflex trilled approximant of
English, the lax retroflex trilled fricative of Mandarin,
the lax alevolar trilled stop of Italian, the tense alveolar
trilled stop of Spanish...-- instead of a flap. Alternatively,
you could simply define /r/ as "rhotic", thus leaving the
door open for it to be pronounced as a flap as well as a
trill, so that those who may experience trouble pronouncing
a trill may use a flap instead (though, of course, the door
would then be open for that r to be mistaken by those
who find a flap to sound more similar to a lateral than to
a trill). Or else, you could drop the contrast and use only
one phoneme /l/ that would comprise both laterals and rhotics
(and rhotic laterals) as allophones.

Second, why having some fricatives defined as voiceless while
some others as voiced? Wouldn't it be more logical to have
a homogeneous row of fricatives? Yes, I know you may have
found v as a somewhat common consonant, but it has to be
considered that the _apparent_ v of many languages is
actually a w that has turned labiodental, as in e.g. Hindi,
Turkish and Samoan, where phonological-structurally their v
doesn't fit in the row of fricatives but in the slot of a
missing w.

Third, it is fairly usual to have the row of stops related
to the row of fricatives as of their number of "articulatory
regions" --I'm consciously avoiding the word "point" here
since e.g. [p] and [f] are not pronounced at the same
articulatory point though in most phoneme charts they belong
to the same "labial articulatory region"--. So, to make
the chart neater, you could either drop the palatal
articulatory region altogether or to fill it completely.
The first option would render a set of obstruent consonants
like this: p/b/f, t/d/s, k/g/h; the second one: p/b/f,
t/d/s, ch/dj/sh*, k/g/h.   (*)suggested orthography: c/j/x

As for the vowels, I suggest that instead of having an
arbitrary set of diphthongs which excludes some fairly common
ones (such as "ei" and "ou") while includes some others not
quite so (e.g. "ia" and "io" as descending diphthongs), just
have w and y as consonants or else allow either all possible
diphthongs or only a "logical" subset of them, e.g. only
those involving i and u as their weak part.

All in all, you could have a phoneme chart similar to this:

Obstruent consonants
voiceless stops: p, t, (c), k
voiced stops: b, d, (j), g
fricatives: f, s, (x), h

Non-obstruent consonants
nasals: m, n
liquids: (r), l
(glides: w, y)

i, e, a, o, u

ei, ai, oi, ui
iu, eu, au, ou)