Hi,
The PV panel shown in post #6 looks as if it is a "real"
crystalline silicon panel, but beware that many cheap "solar garden lights" use
amorphous silicon panels which have several potential "issues":
Firsly, amorphous panels detect mainly visible radiation (light), whilst crystalline panels detect about as much also in the (near) Infra-Red band (as is required, I believe, for a correct watts/m2 value). Amaorphous panels may give a value nearer to a "Lux" measurement.
Secondly, amorphous panels are said to "age" (lose sensitivity) particularly in their first few hundred hours (which may be more of an issue for a product manufacturer than a Home Made system).
Thirdlly, I have measured a few amorphous panels which have had a relatively high reverese leakage current which might affect the accuracy/sensitivity of the measurement.
Finally, for all "PV" type panels, although they are basically a "flat plate collector", there is the potential issue that a high proportion of light at very low angles (to the horizon) will be reflected away by the glass surface and not reach the panel.
Personally, I am using a "normal" silicon photodiode such as the BPW34, which does at least have its sensitivity ("absolute" and "directional") specified in its data sheet. I'm not sure if this is very significant for "watts/square meter" measurements, but it might be significant if one is attempting to determine if "the sun is shining" when it crosses the horizon (at dawn or dusk).
As for the Op-Amp configuration, I prefer using it in a "virtual earth" current-mode configuration (similar to that shown in Appendix A of
this data sheet.
Cheers, Alan.