I disagree. (By the way, Mikuli has D-sharp.) Anyway, the D# makes a B7 (dominant 7) chord which is a natural precursor to an E chord (major or minor, but minor in this case). The D natural would make it a B minor 7 chord.
Terez, at bar 30 there is not a modulation to E minor. Then, D#-G across bar 30-31 would be a really awkward melodic jump. Moreover, notice how long the 7th (A) of the B chord lasts at bar 30 (a quaver) and at bar 32 (a crotchet, compassing the D#). Chopin's writing is full of these subtleties and he wouldn't have spoiled the modulation at bar 32 anticipating the D# two bars earlier.
That is why the same chord (the B dominant 7) is used on the third beat of measure 32 to lead to the E minor chord again. Yeah, measure 29 starts with a C chord. But it is preceded by a G chord, which is the dominant of C, just as B is the dominant of E.
IMVHO you're misinterpreting that passage, there is not any analogy there.
In the attachment you can find a scan of the Henle commentary about that problem. Unfortunately it is in German, so I hope that a charitable German speaker at PS can help us. MindenBlues, maybe?
What I understand is:
1. In the autograph the original alterations are deleted;
2. In the Fontana's copy they are (re)added by pencil, but not in Chopin's hand;
3. All the sources (but Mikuli and Paderewski editions) have D-natural;
4. There are harmonic and melodic reasons that lead to having D-natural at bar 30.