from 62446:
...5.4.5.2 PV string- short circuit test
...test is to verify that there are no major faults within the PV array wiring
~ low irradiance testing cannot verify this as Elecsa explain
...these tests are not to be taken as a measure of module/array performance.
~ results will not be accurate enough for performance testing without taking other things into account ie cell temp./string cable resistance.
...For non-stable irradiance conditions, the following methods can be adopted:
* testing can be delayed
* tests can be done using multiple meters with one meter on a reference string
* an irradiance meter reading may be used to adjust the current readings
~there is no mention of min. irradiance levels but below 200W it is less likely to see current proportional to irradiance (as comments show) and City and Guilds 2399 have included this "200W" threshold level in their corricullum (from what I remember) which is far more relevent when fault finding.
...62446 also has PV string operational tests (5.4.5.3) conducted under normal operation (mpp) conditions, although no reporting of results is included in the model test form.
I also agree that temp. readings are important to performance testing, but here again more so if you are fault finding and need more accurate analysis.
C&G include cell temp.recording and "expected" values for Isc & Voc to be calculated and all shown on their test form, which goes beyond 62446 requirements: although they are covering faultfinding in the course.
~the 5% diversity of string outputs is to verify the design requirements have been met which is more about damage limitation to modules and inverters and was far more relevent when module outputs were more diverse within a batch; on multiple string systems, less so now, modules were moved arround the arrays to different strings to meet the 5% tolerance during commissioning.