Thanks. Yes, before the controller is attached, the array has no current, just an Isc. However, you can measure Voc and Isc and those are affected by shading. But there is no current flowing so how can there be an Isc current bottleneck for a string at the shaded panel?
Once you connect, the controller picks its Vmp and you get an actual flowing Imp, which depends on the wattage. If wattage is fixed, and Voc and Vmp are different, then Isc and Imp will be different amounts. But before you connect there is no wattage because no current?
Anyway, never mind. I am having a chicken and egg which came first problem or something, which I will eventually figure out. No need for anybody else to get tangled up in my confusion. :)