Consider calculating daily aggregates weighted by expected power instead of irradiance

[kandersolar]

Weighting the normalized energy by irradiance when calling aggregation.aggregation_insol is supposed to emphasize values that matter the most for system production. In some cases, irradiance isn't necessarily a great proxy for system production -- for example, a thin-film tracking system can experience reasonably high irradiance during times with significant self-shading loss. In these cases, it may be desirable to weight by expected power instead of irradiance.