The same applies to the corresponding lowercase letters.
It is very much reasonable that the sortName parameter in GlyphData.xml does not necessarily follow the unicode parameter and makes exceptions for compositional glyphs; however, in the case of the above letters, softName probably should follow unicode. It is a bit confusing, at least for a Russian eye.
IIRC, the idea of the sorting was to achieve a sort order relevant for type design, ignoring the alphabetical order. You can still make your own sort order if you feel uncomfortable with the default: http://www.glyphsapp.com/tutorials/roll-your-own-glyph-data
mekkablue, thank you for the quick response. I am not a type designer, but I am curious to know what is so special about the Cyrillic script that makes the current ordering of its alphabet more relevant for type design. Can you please elaborate a bit more or refer to some article where one can read about it?
I was not directly involved in making that order. Generally, Unicode order is not necessarily what makes the most sense in type design or font production. If you are unhappy with the default sort order, you have options: http://www.glyphsapp.com/tutorials/alter-the-glyph-order-in-font-view
“Relevant for type design” means similarly shaped letters should be grouped together. And the order suggested by Ivan does not affect that, since it’s only moving Я to the end.