Regarding measuring accuracy, it is said according to some historical TMA documents that it's better doing a lead-lag maneuver than a lag-lead, which means you can run parallel with the target at the beginning (usually observing a relatively low bearing rate), and then increase the bearing rate later by turning and lagging the target. The goal is to maximise the difference between the approximated 4th bearing and the real 4th bearing.
But it may not place you to a suitable position for attacking.
Speaking of the plotting, I personally likes to go creative using the draw-from-midpoint tool in Wolfpack, like this:
https://www.subsim.com/radioroom/showthread.php?t=240883.