[werk_ag]

So as I see it, the only thing that changes from location to location is the gain needed and the resistances for R1 and R2 to be altered to suit this.

Good to see you studied the matter...
but I think that a sensor must be constructed so as to be universal, and not depend on the location where it will be used, so it is good to build it capable of reading values up to a maximum of 1400 and 1600 W / m2

The only thing that can change the calculations are the solar cell characteristics, it's important to ensure that the R1 value maintain the cell in its linear region. This is what requires quite a few tests.
Unfortunately the solar cell (1V - 100mA) I had used in my prototype units is no more available. Now I'm doing tests with another type of solar cell, more easier to find.
I will give news ...