Take the same 80% for old 3kW system (2400W) there are total of 769 days out of 1980 days where the peak output would exceed the inverters, or 38.8% of days. That's hardly "on occasion".
"On occasion" referred to exceeding 100%, not 80%. I'm at a similar latitudinal situation as you (inverted of course) and I routinely exceed 80%. Nobody is denying that clipping will occur when the threshold is 80%. The question isn't 'how often', it is 'for how much time, and by how much". According to Enphase's data, it amounts to 0.075% of your total energy production. Their paper that is linked above, and in your YouTube video comments, lays it all out, complete with the amount 'lost' due to clipping.
Enphase actually recommends a threshold as low as 75% (133% over-provisioning) and that only because your CEC has mandated that as a limit for some reason. They even explicitly recommend the IQ7+ for your panels over the IQ7A and they explain why. There are, in fact, good reasons to have systems with even higher overprovisioning because maximizing the usage of all available panel power is not the only valid objective, especially since panels have become so cheap. For example, I could replace my panels with new LG ones like yours and new IQ7 inverters that are custom-configured to replace my old ones with a 225W max power output. If you only looked at all the peak power I was 'wasting' due to clipping, it would seem a terrible idea. It would be 165% over-provisioned and my clipping threshold would be 60%, which would horrify some. But when you look at the fact that it would be a simple, easy retrofit that requires no new permitting or infrastructure, it's a great idea and would increase my power production tremendously. Everyone will have their own situation, but these issues are common. Avoiding clipping would lead to poorer system design in many cases, not to mention that SoCal Edison (my power company) really doesn't need a bunch of peak power in the middle of the day.
Some may object to the fact that a system like yours is commonly sold as a '5kW system', when in fact it can never produce that. Up until now I've more or less agreed, but actually I'm not so sure that it is an invalid or even deceptive descriptor. The '5kW' is a sort of sensitivity rating that describes how large a net you have for catching the sun. Whether that is 25m
2 at 20% efficiency or 50m
2 at 10%, it is a valid way of comparing and the peak power issue has such a small effect on total energy production that it really only serves to confuse the issue. IMO, the question is how much your electric bill would change if those inverters were 350W as opposed to 290. If Enphase has given you good data the answer is 'not much'. Then ask what the difference would be if you used 290W panels--the answer to that is 'a lot'.