I'm glad you looked at the two "Melodie" scores--yes, they are night and day! Once someone were to plays the 1940, I believe they would have no interest at all in playing the 1892. I think the revision is a huge improvement. It surprises me though that so few pianists know that it even exists.
Thanks for mentioning the Russian scores site. Yes, I've had that piano.ru/rus in my bookmarks for a few years now. I think the sources of their scores are the same as the IMLSP. For example, in Catoire's "Elegy" for violin and piano, page three is upside down at both websites.
I'll have to compare the two sites for Liadov pieces to see if there are some pieces at one site but not the other.
For Rachmaninoff's Second Sonata I have the Boosey & Hawkes edition with the 1913 and 1931 versions back to back which is really neat for making comparisons.
Rachmaninoff's favorite concerto was his No. 1. He was always disappointed that so few artists played it. The same holds true today unfortunately. I like it a lot. The very long ending is a bit overblown. But aside from that it's masterful. I love the middle lyrical movement--gorgeous! I have a subsection of concerto scores here. Years back, I used to doodle around by playing favorite excerpts from them. But with my recording projects now, I don't have any time for that it seems. No more doodling around!