Usually when I submit it's because I don't see any method for improvement on a solve and assume that I won't get anywhere for a while. I did that last time too, but here I am 15 minutes later.
This breakthrough came by realizing that my method of taking apart the brick in order to let the solution loop actually can be used to remove some latency at the end of the solve because I no longer care about waiting to move the brick when it's convenient for adding on more to the other end. If I just decide that the brick need not exist throughout the entire solve, then I can make use of this to take it apart at the most convenient possible time: when I no longer care about rate and getting more stuff into the pipeline, but only about getting all the products finished.
I'm also now realizing that most of what I'm saying in my solution notes is entirely unhelpful to anyone not familiar with my prior solutions or solution notes. Welp, not my problem