Sometimes it's a good idea to take a step back and rethink the basics. As in, the basics that you've established for yourself in how you've decided to do things. I had a moment of clarity which has resulted in my abandoning the animation class model.
As it turns out, it's easier to take a graphic and break it up, then animate that, instead of taking a mess of body parts and trying to stitch them together into a pre-existing animation. That's how the cowgirl position animations work, but all the showcase, presenting and after sex animations are regular animations.
Animation class sounded good on paper...an artist would provide me with a graphic or vector and I'd apply it to an existing animation. But it hasn't worked well in practice and now-a-days I do most of my own art anyway.
I've created 12 animations so far for v4 and they are all more complicated than any of the v3 ones. Fennekin's and Milotic's animation loops are about 15 seconds long, for example. So I've gotten the hang of doing animations now and I don't find them to be very time consuming, really. It's the graphical prep-work that's time consuming, and prepping for animation classes is the worst.
So I'm doing away with it. Not only will it ease development, but that also means that Liepard, Absol and Delcatty will each have unique missionary animations, instead of them sharing the same one like the eeveelutions do for cowgirl. I've spent a good 4 hours tonight redoing (again) POW's internal structure to work under the new model. Phew!
I've also adjusted the roadmap to reflect the changes. There might be a beta next week :)