Unless there is something I'm missing, the sun still shines from the same direction regardless of the cloud coverage so I'm not sure how having panels pointing in other directions could improve the matter. Perhaps there is a case for optimizing panel area for different times of day but since panels are so relatively cheap it seems the advice is just to get more panels than spend much time worrying about such things.

Are you signing up to point your panels north and take a 30% efficiency hit? Or east/west for a 15% penalty? People point them south because it's the most efficient fixed orientation north of the equator. A more efficient solution is to use a tracker which keeps them pointing directly at the sun as it traverses the sky.

Not every roof allows for perfect southward angling (obviously).

And I'm obviously not saying that you should point panels north either. I'm disputing the parent commenter's claim that it would be beneficial to have all panels aimed directly due south. Because that way you get one strong peak at noon, which is the time of day when solar energy is most abundant but also least used.

The potential for mechanical failures in trackers makes them quite unpopular now (unlike in the 70s when they first started to appear, and seemed like an obvious win).

You're better off just adding however many extra fixed panels you need to make up for the lack of tracking (and its normally not very many).

It might be that south gives you the most electricity (I’m southern hemisphere so north for me), but if you’re after power for yourself, early am and late PM energy generation is very helpful.

A battery helps negate this issue but not entirely.

Pointing west is a reasonable option in California. Pointing west reduces production, but also shifts it later in the day, and addresses some of the duck curve.

This problem has been know for well over a decade...