With just 6 hours left, I’m way behind schedule and I’m not going to make the LD48 deadline.  Here’s what went wrong – hopefully for others (and myself) to learn from:

1. I did the warm-up, but didn’t go far enough.  I really should have gone further and tested every system possible.
2. For this LD48 I wanted to use more robust, and more flexible data management than my previous LD48 entry.  A few of my systems I thought I could build quickly with JS instead of C#… How wrong I was.  Unity’s slight differences to canonical JS became a major pain when I started using more advanced techniques.  When it was time to build on those simple JS frameworks, it just turned out that they just couldn’t do the job.  Ultimately I had to use System.Collection.Generic lists for everything, but I lost about 4 hours overall figuring that out.
3. I opted to use a 3rd party library for sprite management in Unity.  The library worked perfectly, but I just wasn’t familiar enough with it to use it correctly the first time out.  I kept running into what I thought were “issues” — but they weren’t issues at all, it was just my own lack of experience with the library.  LD48 is probably a bad time to be learning how to use a new tool.
4. And finally, (IMHO) the artwork that you need to do a 2D game can actually take significantly more time than a 3D game — it’s more difficult to represent 3D objects in two-dimensional space.  Gradients and shadowing can help, but those are extremely manually-intensive processes.  I’m pretty good with Blender, so in the future I may actually build simple 3D objects and just take snapshots of them from different angles to create my sprites.