In general, at rip time the usual suspects like album, title, track, artist, release year, genre are populated.
The results pretty much depends on the internet database used.
JRiver uses its own YADB and FreeDB
In case of classical, you often find the composer in the artist field and hopefully the performers in the album artist.
Most of the time it is a bit of a mess.
Sonata is MC18+dBpoweramp.
As you have JRiver already, you might consider using dBpoweramp as it pulls its meta data from 4 different sources including GD3