It doesn't prove what you think it proves.
What it proves is that b9 intervals, between certain voices within a
chord voicing, have a very harsh sound and are usually not conducive to
a musical-sounding voicing.
The main exception to this "avoid b9 intervals within chord voicings"
rule is the b9 interval between the root and the b9 in dom7b9 chords,
but there are other exceptions too.
And once you start working with b9's more freely within your chord
voicings you can even develop a taste for chords like your 3 X 3 3 3 7
chord too.
Depends on the style.
Your two chords have the added aural problem that both the maj 3rd and
the min 3rd above the root are included in the chord.
If we hear a min triad in the lower regions of a voicing our ears
usually interpret the entire chord as being a minor chord of some sort.
If we hear a maj 3rd down there we usually hear it as some sort of a
major chord.
When we hear them both together the ear can get confused by the ambiguity.
Since major chords have a much stronger proportional similarity to the
proportions of the intervals within the overtone series of the root of
the chord, major chords are many more times stable than minor chords and
can support all sorts of other extensions above them without losing very
much of their feeling of harmonic stability.
In the theory of acoustical roots that I learned from Gordon Delamont's
book it's said that #9's above major chords are experienced as distorted
versions of the 9th partial.
I.e. The ear hears a harmonically strong and harmonically unambiguous
major chord on the bottom that closely mirrors the OTS, so the only way
it can make sense of the #9 on top is to hear it as a distorted version
of the 9th partial.
Minor triads, on the other hand, don't have the same type of harmonic
stability as major triads because each interval within the minor triad
has a different acoustical root.
E.g. In a Gm triad, the root of G-Bb is actually Eb.
The root of Bb-D is Bb.
However, the root of G-D is G and when all 3 notes are played
simultaneously it is the harmonic strength of the G-D interval that
governs our experience telling us that G is the overall root of the
entire chord.
Delamont says, and I agree with him obviously, that the min 3rds within
minor triads are experienced as distorted versions of the 5th partial of
the root.
So, if we have a minor triad on the bottom of a chord voicing and we
then tack a major 3rd above that, the ear gets very confused about the
overall harmonicity of the entire chord because there is an altered 5th
partial in the lower regions of the voicing and an octave double of an
unaltered 5th partial on the top of the voicing.
I.e. Adding A B natural above a Gm7 chord makes us hear it as a really
bad voicing of G7.
If your original chord-with-the-extension-on-top had been a G9 chord
(voiced G F B D A), and you switched the positions of the B and the A
(to G F A D B), there'd be no problem.
Of course that particular voicing of G9 isn't really physically possible
on guitar in standard tuning.
Try it on piano though and hear that it sounds fine and still sounds
like G9.
But 9ths are voiced below 3rds all the time in all sorts of chords.
And 11ths can be voiced below 5ths.
And 13ths can be voiced below 7th.
Success depends on the specifics of the musical context and the
specifics of the actual voicing and on taste of course.