In this case, my goal wasn’t to test the PR (which had already been merged for over a week at the time I updated bob).
The goal is more: How does this change perform against what’s currently released and deployed on the network. While having another one or two nodes with this setup would have been nice for a bit more confidence in the data, running it on all other nodes would have made it impossible to compare old vs new.
As part of my 30.0rc1 testing, the nodes now all run with the defaults decreased by 33106.