I called myself done with this puzzle after spending at least 12 hours making a 4 mechA solve.  That was a very painful ordeal.  I didn't believe 3 was actually doable, since I didn't see a way to exchange positions of pieces in their order around the arm.  With 4, you could build a C and drop it off on the left grab point, pick it up on the right grab point.  So I gutted out a 4, put the puzzle down, and tried to walk away.  I said I was done.

However.  Broken promises.  I came back on Friday to try 3 just a little.  And wow it wasn't much harder than 4.  In fact, this has far better cycles (and area) than what I had made previously.  I did learn a lot from the previous attempts.  I learned about the specific ways I need to consider bricks and vectors.  I sampled every flavor of syrup under the sun.  I was much more aware of the big picture so wasn't hasty placing glyphs in the near field.  The only glyph in the same zip code as the arm is the one that fires first.

Thanks to the insights from the painful 4, this took one sitting of around 6 hours.  Two of those hours were spent entirely on just rescuing the waste chain after output - I found a way that added around 23 cycles per tape loop to make a reachable handle.  Then I 6Pd it to save those 23 cycles on the last loop.  If I am overpromising effort to this week, I gotta full commit.

Now I'm done.