Circling back, yup K 0.08 makes a big difference using the test generator sample.
However, has helped for prints that have small gaps, I had a problem that they would always end up welding together.
My default print speed is 60mm/s
Nice improvement lowering the K-factor for fine work. Notice the gap in the middle is clear. Ignore the quality of the second print, turns out K 0.10 was best for me but I took these at 0.08.
Also can effect retraction, before:
After: