Help needed choosing MPPT charge controller

[Hedges]

You mileage may vary.
On higher latitudes, during the winter, when sun is scarce and you have pain to harvest a few Watt, you can't afford to lose mA into a shaded panel.
The sun is also heavily turning so you really need the SE and SW orientations to catch every short sunshine that could pierce the clouds.

Pay me milliamps, or pay me volts.

What is your string Vmp under those conditions? What diode do you use? What is Vf of the blocking diode at that current?
Do you actually get any more power by using a blocking diode?

Especially when there are clouds, there is a glow falling on the other panels too. But anyway, at Vmp there is very little leakage current. Voc is about 20% above Vmp, and current through a diode (the solar cell) is an exponential function of voltage. That refers to the current leaking through the forward-biased PV cell, so not available to your charge controller.

Analyzing the graph on PV panel data sheet, or curve-fitting to it in Excel, should provide the answer. Field measurements are less dependable, but a lab setup could work. For instance, when I did the experiment of shading one panel out of 9S2P, variations over time were about as much as the data I got - not a perfectly clear day.

 

[rin67630]

Especially when there are clouds, there is a glow falling on the other panels too.

You did not catch the point: with casted clouds, it does not matter at all, you get 5-7% of the rated power at the best, with and without blocking diode.
But sometimes you get a sun ray for an hour or two and that's the chance to harvest more from that time than from the rest of the cloudy day.
Provided you have a panel that is well oriented at that time, because the sun turns much more on higher latitudes.

The diode will be a(several) schottky diode(s) rated a bit higher than Imax of each panel, obviously.