Having point keys on every frame was a BONUS for me. I don't really need it.
What I initially needed was only ONE point mesh key frame for each key of the bone layer, just as you did in your "new" version. And also I got that with the original layer scripts.
I do not need keys on every frame. I am not worried in the least about inbetweens not "matching" the bone rotations. I never have that problem when doing switch layers anyway and they use point to point interpolation. If I get my points on one frame I'm happy.
I could easily add a bone key frame ANYWHERE between bone keys in the animation if I needed to and convert that to point motion!!! Bingo!
You see, it doesn't matter to me how many kludgy hoops I have to jump through. I have the ability now to turn ANY BONE MOTION into VECTOR MOTION. It doesn't make any difference how "hard" it is.
From my point of view, you have no idea how cool this is for me. It is just... cool as freaking heck.
For instance, I have larry the Liger. I was trying to produce different expressions. I would have to draw them all. I would have to modify point by point by hand to produce new facial and head distortions for switch layers and lip syncing.
NOW I can use BONES WITH CONSTRAINTS to distort his mouth, his head, squash and stretch very easily and very quickly without messing up the original mesh. If I don't like his snout in that position, just rotate the bone.
I convert that to point motion and I have beautiful wonderful switch layers for use with lip sync and papagayo.
Now of course... others may want more features. I have what I want.

I see this script as a way to take vector "snapshots" of bone motion.
<sigh>
I'm getting light headed again... I better stop.

-vern