The ghost of Sheldon Brown is screaming "The perfect is the enemy of the good." If I remember right, in a motorcycle MOT you are allowed 2 mm wobble, and remember that motorcycle achieve some pretty high speeds. I would guess that the pinging technique is good enough.
I would also say that wheel building follows nature`s 80 / 20 rule. It will take 20% of the time to achieve only 3 mm wobble, and 80% to get it down to 2 mm. Whatever method you use to measure tension, always remember that in a dished wheel, the tension will be higher on the drive side.
Also consider this. Think of an engine cylinder head. To ensure a gas tight seal, all the head bolts need to be tightened to the same tension. How do we measure this? We use a torque wrench. So why are we not using a miniature torque wrench to check our spoke tension?
Here is another theory. If we have a perfect hub, a perfect rim and perfect spokes, we could achieve a perfectly true wheel with all the spokes at the same tension. All these components are mass produced. Variations are inevitable. How round is that rim, to start with. If you placed it on a surface plate, would it touch all the way round? Is the tensile strength of the metal uniform throughout? I doubt it.
Therefore, to achieve a perfectly round true wheel, the spokes will need slightly different tensions to hold it in shape. We are looking at a good old British compromise. Part science, yes, but importantly part art.
