Not sure if there are any plans to incorporate ninth chords (dominant, major, and minor) into Hookpad, but it would be a nice touch. Right now they can be approximated by 7sus2 (or by a 7th chord with the 9th in the melody, which is sometimes possible), but having the real thing would be helpful.
Unlike most of the currently unsupported chord extensions, it wouldn’t require any changes to the interface - you could just allow the 7 and add9 boxes to be checked simultaneously.
How then would you expect the ninth chords to behave? Hookpad has never supported five-note chords, and what you are suggesting would probably use the omitted fifth voicing as suggested by most music theory books. A lot of such chords will remain ambiguous unless otherwise stated in the comments, although this may lead to more niche uses of the ninth chords represented in Hookpad due to the omitted chord factors, for example:
iii7sus2 becomes 7♭9
viiø7sus2 becomes 7♭9♭5
vi9no5 becomes m9♭5
V9no5 becomes 9♭5 or +9
I9no5 becomes +M9
These non-standard interpretations will come into play as long as Hookpad continues to employ four-voice chords; on the other hand, if ninth chords are added to Hookpad without any omitted chord factor, they would become handy if one wants to express them correctly, but ultimately they won’t add new choices of chords because all 7sus2 chords already represent the respective ninth chords as long as the chord data supported by Hookpad remain diatonic.
I’m not nearly as familiar with Hookpad as you are, and hadn’t thought through all the implications of four-note vs. five-note voicing. Given the factors you mention, omitting the third (as in 7sus2) seems at least as good a substitute as omitting the fifth. As for the prospect of a full five-note ninth chord, though, I do think expressing chords as accurately as possible is a worthy goal, as a 9th chord is meaningfully different from 7sus2, and that difference is worth expressing (if it’s feasible to do so). Certainly adding ninth chords wouldn’t be as significant a change as adding chords for which there’s no satisfying approximation (e.g. augmented), but the interface at least could remain the same. Far from a big deal, but it would be a rather nice feature.
It would actually confuse those who know music theory instead; add9 implies no7, similarly add11 implies no9 so it won’t be as simple as permitting “7add9” in the interface.
The greatest barrier would still be Hookpad’s algorithmic voicing. In an extended chord the extended factors (in this case the ninth) must be more than an octave apart from the bass, and they almost always appear at the top. However, Hookpad might consider, say CM9,
a particular inversion of Em7 on the right hand over C on the left hand, creating a Em7/C chord during run-time. The inversion of Em7 is necessary, otherwise the same voicing issue as the current sus42 chord will happen; or
the exact transcription of CM7(add9), handling the D note separately and forcing the extended factors to play sound an octave higher than usual, which might incidentally clash with the melody track due to high register and volume; or
a C5 power chord on the left hand and Em7(no3) on the right hand, but such voicing is of course inconsistent with the four-voice chords, and this would make extended chords sound thicker than the other chords.
True; for Hookpad users thinking of add9 as an operation applied to a chord, which I suspect is a lot, “7add9” would be completely intuitive, but the strictly correct view of add9 does make “7add9” a contradiction.