Accuracy of PV and/or Wind modeling?

Hello Paul,
Can you give me an idea of the accuracy of the output of SAM?
Let me be more specific:
1. If I build a "PV power plant" at location "A".
2. I then do a SAM run of the precise same plant in location "B".
Taking ONLY the power generated, would the kWhrs/year be accurate within 1%, 2%, 5% or x?

I assume the variables are:
- The accuracy of the pyranometer used.
- The accuracy of the data notator whose records get loaded into the SAM weather file.
- The accuracy of the PV module insolation/power output curve.
- etc.

I understand that weather changes each year, but that is not my question.

I am trying to understand the case where if location "B" indicates 2% more power generated, is that site REALLY likely to demonstrate 2% better performance over time?

Thanks,
Tom