What I do is visualize a gaussian distribution and presume that WGT can deviate me up to 4 sigma. That's a rare deviation, but it is gonna happen (and it's large, when it happens).
I have these figures that I use for lie and spin and wind. I developed those early and they were accurate. I don't know if they became inaccurate because I advanced in tiers or because WGT toughened up the deviations. It doesn't matter. I trust them. If they don't work, it's because of deviations.
It's a hard row to hoe. I used to make 20 yard adjustments on St. Andrews and be confident I was going to be within a couple of yards. Now the wind may drop off entirely and I'm 20 yards off. Most of the time, though, I will either be right on or only 3 or 4 yards off what I figured.
What I do is set the triangle where it should be according to my trusted figures. Then I look at it. I ask myself what's going to happen if the effects are zero or doubled, and what is going to happen in those cases. There's usually some bad trouble in there somewhere. Then I readjust my aim point. I'll try to pick something that keeps me out of trouble, even if I come up a dozen or so yards off. I'd rather putt or flop a bad shot than be in the water or out of bounds or in lebenty-jillion % fescue or hit a dam' tree.
If there isn't any bad trouble, I whang away at the pin (unless there's known roll to the right or left, in which case I whang away at my predetermined landing spot).