Well, that only pertains to my arrangement of Meteo for the album. HOWEVER, in the original, the "da-ga-da, DA" accents happen on beat three of each measure (all 3/4 meter), which threw me off. It took me a couple listens to figure it out, so I decided to play on it, which gave birth to the current WIP that you have.
In my WIP, which I can't really link here (very sorry), I keep the A sections in 4+3, with the final bars of it in 5. The B section is in 5+6 (so I actually compressed the melody by 1 beat). I have nearly decided that the C section will be in straight, easy 4/4.
Soooo thats a rough metric breakdown of my WIP, I guess. Its both easier and harder to arrange source material that has static harmony throughout whole sections. I like playing with time and groove, but it takes a lot for me to inject more creativity on top of that, which unfortunately shows my hand in terms of arranging.
In terms of key changes - I was INCORRECT in my last post
What actually happens is, lets say the tune starts out in E minor. It modulates to EITHER Eb Major, OR Ab Lydian (those two have the same collection of pitches).
If it goes to Eb Major, then the chord changes during the B section are ||: IV - V - IV - V :|| or "AbMajor - BbMajor"
If it goes to Ab Lydian (has a raised 4th scale degree), then the chord changes are ||: I - II - I - II :|| or "AbMajor - BbMajor"
In either case, the chords are completely diatonic which is pretty cool.
And truth be told, I REALLY would like a second opinion on that analysis.