In this sense, I agree with Schoenberg: the chord is an expression of a scale only on a theoretical level, not on a practical one. And in fact, I must say that I have never come across harmony treatises stating that the Neapolitan sixth is a direct reference to the Neapolitan scale. If it were, we would have to "postulate" a sort of modulation every time it is used. And a significant modulation at that, since we are talking about a minor key. But this makes no sense; we are even talking about a major chord, which, if inserted into a cadence as usually happens, gives me no reference to a hypothetical minor scale, whether Neapolitan or otherwise.
It is a problem that, by training, I would be inclined to resolve historically. But I have no information regarding this; I have never dealt with Neapolitan music, but perhaps in the 1600s there actually was a Neapolitan minor scale upon which this chord was normally built, and since Neapolitan music traveled the world, everyone else in Europe knew of its existence.
But if we postulate the existence and widespread knowledge of this scale, then we cannot limit ourselves to taking just one chord from it... in addition to a major chord on the II degree, we would have a major chord with a diminished fifth on the V degree (!) and a diminished triad with a diminished third on the VII (!!). Now, since we are talking about significant collateral events—that is, two of the most important chords that drastically change their own nature—in my opinion, the Neapolitan scale is not a true tonal scale. Perhaps it was used only for melodic purposes, or rather, THAT second degree there was used only for melodic purposes and supported by chords on the II. But regardless of the fact that anything is possible in history, I have strong doubts that dominant chords constructed that way could have been used intensively. I mean, a diminished fifth on the V?! In a period when people were still "afraid" of discovered fourths?
I continue to think that things went differently (but these are just suppositions lacking historical data, alas):
1) a normal Phrygian cadence, in the modal sense, starting from a minor sixth dyad (let's assume F-Db) can resolve to a fourth dyad (G-C), precisely because there would be the Phrygian tenorizans clause (Db-C) and the Phrygian cantizans clause (F-G);
2) the fundamental "core" of a Neapolitan sixth chord is its minor sixth, since it is always in first inversion, and its "traditional" resolution is to the second inversion of the tonic chord, which has the discovered fourth;
3) it is probable, then, that one day someone wanted to replicate the effect of the Phrygian cadence but by transporting it into a tonal situation, and thus did so with inverted triads;
4) the choice of the major fifth was mandatory, obviously, otherwise it would have become an augmented triad;
5) and when one day someone asked "but how do we justify this chord in tonal theory?"—and they likely asked this very, very late, since Rameau calmly admitted chords with an added sixth (though, if I'm not mistaken, he only considered major sixths)—then they said "let's build it on the second degree of the Neapolitan scale and say that it must obligatorily be in first inversion."
A bit of a complicated explanation, but I hope I have hit the mark...