As a reminder:
1. Squishing textures together would make the creating process incredibly more difficult and time consuming for us, and is Right Out. We'd have to have some manner of determining where frames are via some new process, or define it all by hand, and that's not something I'm willing to invest that level of time into.
2. Being able to switch to 16bit textures is not a possibility with the unity engine the way we use it. Their flexibility when it comes to importing at runtime is minimal at best, and our pipeline (for reasons already outlined) absolutely depends on being able to import at runtime.
3. The people who aren't of an opinion that the art is "a little bad" or "in need of slight tweaking" are likely not going to be looking at anything like what you mention. We deliberately and painstakingly pick the best of the best screenshots to show with press releases (as do all games), and people still kvetch over them to an extreme degree. IE, to reach those people an entire overhaul is needed.
4. Whether or not a kickstarter is theorized to work or not is irrelevant; we're going to try it. If it doesn't work we've learned something interesting from that, and if it does work then we've got a new art style that more people like. I'm not investing money into prototype art at the moment just to back away at the last second because a kickstarter might not work, you know?
5. Yes, without the post-processing the rendered 3D stuff at 128x128 or even larger looks abominable. Blurry, no contrast, etc, etc, etc. Hence our move to drawn art, since the post-processing approach has led to so much divide. Also, it would be nice to have people with more formal expertise in art (plus their "10,000 hours" in) compared to myself, a relative amateur when it comes to art.