Greetings to everyone.
In general, there are some characteristic sequences in Fauré (some also in other authors) related to these themes:
- harmonic connection via common note/s with possible defunctionalization of the individual chords
- use of chords with an added sixth (especially on the IV)
- V-II and II-IV relationships (not so common in other authors)
- use of chords deriving from contrapuntal motions (perhaps chromatic)
- use of augmented triads with a "suspensive" meaning
- characteristic cadence (I had started a discussion on the old forum that ended badly...) and particular interest in general towards plagal cadences, also with variants in both state and added/chromatic notes.
In the work in question, compared to what AleGozzo has already indicated, I would see:
- mm. 7-10, a transitory modulation to III (si-) with harmonies of VI=IV6 (m. 7), I46, IV with an added sixth that makes a "softened Fauré cadence" (the subtonic is missing), I with a 6-5 appoggiatura which then returns to the V of G
- mm. 11-13, an inner line G-F natural-E that does not cause modulations, only color
- at m. 19 the change of mode using the I-III connection (seen in the context of F major as II-IV), which then with the added sixth at m. 22 truly becomes the IV of F major with a consequent I46 at m. 23; the subsequent C-C# chromaticism and relative augmented triad leads back towards the V of G
- following the deceptive cadence at m. 27, there is a moment of "plagal bass" that leads to mm. 31-32 on the common G which links the chords well indicated by AleGozzo