Movements one, three, and four exhibit Phase B piano playing and certain moments of the Sonata in G minor remain unsatisfactory.
In the first movement, a harmonic structure was planned for the exposition (from [0:00–2:11] and then repeated from [2:11–3:52]) and recapitulation (from [5:05] to the end of the performance), but the development section (from [3:53–5:05]) unfolded without any specific harmonic plan. The repetition of the exposition section is different from the first, because no written materials were used during the performance. A small critique can be made about [3:00]: a modulation going straight from an augmented sixth chord to the dominant without passing by the tonic in second inversion is unusual.
The second movement sounds satisfactory from beginning to end, but also somewhat harmonically repetitive and unadventurous. The diminution of the inner voices at [3:05] creates some sense of development, but the harmonic content remains within the same territory for the duration of the performance.
Because the third movement was composed and then realized in performance, it exhibits both Phase C and D activity.
In the fourth movement, mistakes start happening from around [0:32–0:44] and again from [0:59–1:14]. And, in general, the tonal characteristics of the various contrasting sections to the rondo theme are not harmonically specific enough – each of the contrasting sections enters many different harmonic territories, rather than one specific section exploring one specific territory (for example, the first contrasting section could have explored the relative minor and the second the subdominant). Harmonic material that I tried to reserve for the coda, namely the harmonization of a descending bass line, already enters around [2:55], with still over two minutes of playing left. And finally, the whole five-minute movement seems to be performed at one general mezzo forte dynamic.