You cannot accurately model a solar cell with these two components.
So you must decide what is important, and your model must model that accurately.
Your own answer is (was) correct as far as it goes, and will accurately model the cell at a single point, the MPP point (and sufficiently close to it) but will be inaccurate at both extremes of open circuit and short circuit.
As you are modelling a simple curve with a linear approximation, you can intersect the curve at 2 points, providing 2 accurate answers. Three obvious choices are:
- correct at open and short circuit points, which will woefully underestimate the cell's performance at MPP
- correct at short circuit and MPP, which will woefully overestimate the open circuit voltage
- correct at open circuit and MPP, which will woefully overestimate the short circuit current.
The last of these is probably most useful for most purposes; the Thevenin calculations should be easy. (However it should be used with caution for simulating MPP algorithms because the gradient is wrong; for that purpose, your original answer is better)
Open circuit and MPP are important conditions for e.g. solar inverter design, performance at MPP, and overvoltage protection, hence my choice of the third option.
However, for testing short circuit protection, I would switch to the first option, so I would calculate both of these. (The second option would give you false results on overvoltage protection circuits, I can't see a good reason to ever use it)
