2oldman: my 2*24V panels are in series and have 4 bypass diodes each. I didn't do extensive shade testing, but when total voltage dropped from 60 to 30 (which means one panel is out or almost out), the amps dropped a lot. Though with only 2 panels, when one is affected a lot, the amps are expected to drop a lot. Once or twice I saw the dark panel dragging the total output close to what one bright panel alone would deliver, if left alone.
In a little shade testing that I did on a single panel, the output of one panel dropped proportionally to the shaded area, but at some moment the output dropped completely while the shade was between 70% and 80% of panel.
Now when I think about it, if one of 2 panels is affected so much that its "almost out" - say, more than 75% of area is shaded and/or the shade extends over all 4 strings - the total output would drop in half of course.