Notebook temperature coefficients could be improved

[kandersolar]

The PVDAQ4 notebooks assume a temperature coefficient of -0.005 1/K. This might be undesirable in two ways:

• -0.005 1/K is rather large in magnitude; a smaller value would probably be more generally suitable if people are leaving it at the default when they swap in their own time series data
• -0.005 1/K doesn't match the datasheet value for this system's modules, so it's not correct for the particular system under test either

After offline discussion it was decided to add an explanatory comment in the notebooks for now, and in the longer term (RdTools 3.0) update the value to something more suitable.

For reference, it seems that the -0.005 1/K value has been carted around ever since 2017 when the notebook was first added (see #17 and specifically 161dec5).

Thanks @adriesse for pointing this out!

[adriesse]

Thanks ADR PV efficiency model for detecting the correct value in the data! :)