A slightly optimized area solution. I think at least five of the eight theoretically-removable area is actually possible to remove, so we should see primaries of 25 or less.

The layout has enough space for a second arm, which would probably make it faster, but I didn't want to have to do all the piece juggling to actually test whether it outweighs the resulting restrictions on track movement. As it is, this is more of an AGC solve.