Mixed panel orientation array

Here is a question I am almost embarrassed to ask. I think I know the answer,, at least intuitively but something niggles at me.

I have two sets of panels, one set of Siemens 55 with a Vmp of ~17.4 vdc mounted on the wall. I have another pair of BP 63, also with a Vmp of ~17.4 vdc. mounted on the roof. Each parallel pair, feed the CC with a pair of #4 wire, combining at the controller. Now I know that the Vmp of the entire string is regulated to a great extent by the LOWEST Vmp in the entire string.

I also understand, that if in the case that any panels (or parts of panels) that are wired in series, get shaded, that shading will have the effect of reducing the power out of the entire series string.

My wall mount panels get sun early, but loose sun late. The roof mounts do the opposite. I am going to add a 5th panel, bringing my total capacity up to about the maximum that the controller can handle. (More than I really need,,, but you all know how that goes!) This panel will be a rover, set up so I can move it both during the day, and over the course of the year. It will probably feed through the wall mount wires.

Now my question: It is my believe that shading of any one panel will not effect the output of the entire string, only the effected panel. Is this assumption correct? My concern is the 1/4 of the days when one set or the other is shaded. I monitor my amperage numbers pretty carefully and it seems to prove my supposition, but usually by the time the wall panels go in shade, the MPPT controller is dialing down the amperage anyway. So the long and short of it is,,, should I set all the panels up to get ~ equal sun at the same time?

Tony