Hi gruffalo,
Sorry. Back to business.
I remember listening to your first rendition, but just now listened to it again before going on to the the second recording, so heard them both today. (I've not studied Gaspard, have just read through parts of it.) I've heard the recordings of Gieseking, Lorti, and Ashkenazy among others, so I certainly know the sound of the piece extremely well.
Overall, I like your first effort better than the second. The tempo is slower in your first Ondine, but that's OK. If you haven't watched Perlemuter play it on YouTube, tune into it. Perlemuter, as you know, studied Gaspard with Ravel himself. So as he played the piece throughout his career, he did so with knowledge that no other pianist could have. Perlemuter's video was made when he was in his 90s, so they are some wrong or dropped notes once in awhile. But it's like watching history. Notice that he doesn't play Ondine like a speed demon either, but nonetheless he achieves astonishing beauty in his performance. That's why I take no issue with your slower tempo there. This, after all, is Ondine trying to seduce the poet Bertrand with her cloying, pleading, melancholy song. We think of this piece as impressionistic, and it is, especially the shimmering effects of water. But it also has romance in it too. For it to be beautiful, it has to be somewhat gentle and leisurely. At that tempo, I think you give much attention to detail as well as playing the long line. The dynamic control is excellent as are the shading and nuances. The coda could have been played a bit better I think. But overall, this is a good, creditable performance. You really put it across to the listener.
The quality of recording of the second attempt is not as good in my opinion. Also the tonal balance between the hands is not as carefully controlled as in the original. In fact, background sometimes battles with foreground. Overall, it is faster and played with more confidence, but it's also more strident at times too.
Sometimes when we first undertake a piece, it contains a huge measure of inspiration. That inspiration overcomes the few evident fluffs and flaws. Later, when we decide to revive and refine the piece technically, there evolves a more marked correctness to it... but we cannot always recapture that inspiration in exactly the same way as before, which might be the case here. There are certainly improvements. But if I were you, I'd still submit No. 1. Good luck on that!