The problem is that the maximum power point is constantly changing depending on how the light is hitting the panel. And the voltage can be radically different when you have a shadow on the panel.
A typical MPPT hunt will start at the full open circuit voltage, and apply a small current, then measure the resulting watts. It then increase the current a tiny bit. Did the wattage increase? If yes, then increase the current a tiny bit more. Did the wattage increase? If yes, increase current again. Keep repeating until the wattage falls, then reverse. Lower the current. Did the wattage increase? IF it did, lower current again. If the wattage drops, revers again, and increase the current.
The routine works just fine if the amount of light changes slow, and you have no shadows. But when something moves and a shadow appears on the panel, the maximum power voltage is going to change a lot. And it can take a while for it to find the new power point. And with a small shadow, the maximum power curve might have a second knee point where the bypass diodes take over and the current can increase a lot more at a lower voltage with a column of the panel that has a shadow being bypassed. Many MPPT controllers will only find the higher voltage knee, not a lower one. And as the complain here shows, a simple MPPT search can go bad and run to maximum current and drag the voltage down. The Victron way of resetting every 15 minutes and doing a faster hunt might make more power in tricky conditions. But it might make less power over time if the light is steady and only changing with the earth rotation. A slow tight search will stay closer to the max power point while the reset totally moves out of it and has to find it again.
Midnight Solar charge controllers have a few choices for the search routine. But it could take many days to find what works best for your panels and location. My cheapo BougeRV charge controller seems to search fairly fast, but I do not se it doing any reset like the Victron. But I do see it swing over 50 watts past the max power point, both up and down in voltage, about every minute or two. And that is at 700 to 1,400 watts with 2,000 watts of STC rated panels. In reality, that 50 watt swing is small, just 7% at 700 watts, and less than 4% at 1,400 watts. And when it drops by 50 watts, the next step is back up in wattage to the peek again. Over time, they claim it average at better than 98% of the true maximum power. I don't know how true that really is. On the good side, the only time I see it pull the panels down to battery voltage is when the sun is very low in the sky.
I would love to try a Victron on half my array and compare them for a few days and see which really does make more power and by how much. Are they worth the cost difference?