OK, so if you lay ground panels flat like roof mount, then you can compare readings on two SCC.
That might flag any issues in panels or wiring.
What I found first was one paralleled string delivering zero. A wire nut had slipped off. I'm upgrading to Ideal setscrew nuts as I go.
And then I also found the bad Sharps.
Most of my panels go through touch-safe fuse holders, so by shutting off GT inverter I can open some, then turn on and wait 5 minutes for a measurement. But the sharps are paralleled at the array, through a larger gauge wire, so unplug MC connector. Or use clamp ammeter for non-invasive test.
If there was a non-contact voltage measurement at MC connectors, could diagnose to panel level. For car wiring, insulation piercing is done. With a pile of MC "Y" cables, could add test points. I knocked down the series connections and tested individually, finding bad ones. But lost track of my interleaved panel order, so couldn't tell for sure if the bad panels were possibly biased the most negative (or positive) during their year of living transformerless. "PID" might be an issue for them negative biased, vs. always positive biased for their first 17 years of life. Second string in parallel unaffected or not severely. Dirt and moisture figures into environment that causes "PID" as well as other mechanisms.