I don't like those NOCT figures much, because in real life, here in the Spring I would be getting full panel current while panel and ambient temps were quite cool.
They try to connect lower temp of 20C vs 25C with lower irradiance than STC. Doesn't happen that way. Separate events.
(Never mind that to get panel at 25C, ambient would have to be near freezing !)
I did measure the lower panel watts/voltage (but not current which rises) with higher panel temps. The graphs that show 10% loss from rated at 25C when panel is 50C comes out right when I checked that. (Using my IR gun on back of panel with panel facing the sun)
Looking at the Vmp of 57 in the OP, with the string partly shaded, (it is the chosen Vmp as seen by the MPPT being in Bulk) if you use all six, 57/6 = 9.5v But if two are by-passed as Mr Wiz figures, then 57/4 = 14.25 as average panel Vmp.
So it seems if the string is 4 in the shade and 6 in the sun, the irradiance is such that chosen Vmp is 14ish per panel either way.