I wonder if this would sound better in a string quintet setting than the full orchestral strings. With the GPO4 solo string patches, I like how it sounds. I'm imagining in the context of a bit of pedal use by the piano player.
As for the notes, the Bb and the B-nature got moved to the viola part and now are on octave lower than they were in the original. When I'm a bit more awake, I don't necessarily agree with this voice leading.
I believe what I was trying to do, with both the Vln II and the viola part ending on G in the first measure, I didn't want parallel movement by them both going to C, so I chose the viola to go to C and the Vln II to go to Ab as a way to emphasize the fact that this isn't a standard chord in C major any more, the same reason, I doubled the G# in the third measure with the double bass harmonic an octave below.
Here is another treatment of the same with a bit different soprano line. I have to say I prefer the harmonic structure of the first way first, but to each his own as to what the ear hears.
Caleb, again, the problem is that you are breaking one of the conditions of the exercise.
By altering the soprano line, dropping part of it one octave, you are breaking the melodic contour and creating a "melody" that does not exist. While this might work in a development section, to have a melodic contour skip octaves between different instruments, in the strict context of orchestrating a passage, it doesn't work.
Your first example drops the last two notes of the melodic contour, while your 2nd example adds a new high C in the 1st violin part.
Again, twice you break the contour of the original.
There are times where tying to be inventive can also be self-defeating.
In this particular case, the most effective answer to the question is not necessarily the most complex.
