There is a danger in listening too often to the same piece while it is in progress or immediately after it's completed. The closer we are to our music, the more inclined we are to like it simply because it's familiar. One of the things I've built into my own composition process is a waiting period after I complete the first draft of a piece to gain a measure of objectivity on it before beginning to edit.
Regarding the ease of the harmonic transition between into m. 24, half diminished 7th chords are more dissonant than minor 7th chords. So, despite being the correct function for the spot, it will not make the transition sound any easier. Your progression immediately before also implies at least three different keys before landing the listener in D major. It's a very nicely written piece of harmonic deception, by the way.
Now, to the more philosophical subjects...
On the subject of recapitulations, development and recapitulation is different in concept from simple ABA form because development has consequences.
In the simplest sonata form, the development features enough harmonic instability to solidify the tonic key such that both subjects are stated in the tonic key instead of in separate keys. The first movement of Eine Kleine Nachtmusik is a good example of this. The development does not go terribly far afield form the thematic material and the main subjects is restated nearly verbatim while the second subject is modified to work in the home key.
In more complex sonata forms, the creative ideas from the development actually change the nature of the recapitulation. In another Mozart work, Symphony No. 40 in G minor, the development section is designed to fake the listener into thinking they're hearing the recapitulation. The entire development is about deception. So what does Mozart do when it's time for the actual recap? He brings the violins ahead of the lower strings, making the real recap look like yet another fake. It's really hip writing. The recap does its job of returning the listen "home" musically, but it carries forward the idea of the development.
Beethoven, being more strongly driven by expressive desires over form, was more aggressive in bending recapitulations to fit the narrative context. In his 5th Symphony, the recap of the first movement is orchestrated very differently from its original appearance, carrying over elements from the development in the background. The orchestration is lighter and perhaps even more "tired". There is a reflective little soliloquy in the oboe at one point. These changes feel like a consequence of the development, which portrays a person weakening the face of fate.
In the fourth movement, Beethoven wants the recap of the A theme to feel very much the same as it did the first time, like a bright, victorious fanfare bringing light to the darkness. So, not only does he recap it verbatim, he precedes it by quoting the last portion of the third movement so the fanfare has the same dark, foggy texture to contrast against.
Structurally, it boils down to this: the bigger the difference between the subject and its development, the bigger the consequence that has to be dealt with in the recapitulation. Keep that in mind as you're studying classical works.
On the question of making music for the listeners vs. you personally, it's a balancing act every composer has to grapple with. The challenge of communicating through music is finding the right metaphors and language to make ourselves understood to our desired audience. Sometimes, the thing that totally makes sense to us as composers doesn't work for an audience. The question then becomes whether you as the composer like that thing enough to stick to it or whether it makes sense look for another solution. It's a judgement call each and every time.
When it comes to finding out what works and what doesn't, this forum is a great resource. Not only will folks tell you, but they're fellow composers and can tell you why.
Finally, nothing says you have to go back and revise this piece. I believe it was Tom Petty who said that projects are never completed, only abandoned. We grow as composers by tackling new ideas and writing new pieces.