New metric, who dis?

*********

My interpretation of this novel metric applied to this puzzle is that *glyph throughput* is the name of the game.  In pursuit of that, I tried to arrange a layout that facilitates a simple clockwise flow of atoms, where the calcifier and duplicator are easily accessible from both the input and the triplex bonder.

And it works, more or less.  This machine produces all of the firework outputs at 5r and explosives at 4r.  I consider it only a partial success, though, because 4r for everything is staring us square in the face; in each of the three cases, the retracting catalyst blocks either pulling the next input or swinging fire off the debonder.  A problem so small it seems veritably certain that someone else will find a way around it.  If not via superior layout, than some method of batching fire-then-elements which seems entirely too annoying for me to contemplate.  I'll stick to my small-brained strategies, pleaseandthankyou.

**********

One of my proudest achievements here was finding a way to swap catalysts with only a 3-cycle delay.  As a result, I think my latency is quite good; small-brain strategy or not, I feel confident that this is a strong implementation of it. If folks can unlock big-brained methods for 4r or less across the board, I expect the top solves to clock in under 110c. 

**********

Oh, and about the solve name.  When I was debugging the last products, my wife commented that the sound of the mechanism during the construction of Explosives reminded her of the Capybara Song.  

Triplex music is real!