You can attempt to re-upload new settings to the mouse over and over and sometimes it will get close to how it was from the factory, but generally never quite as good. It's more difficult to notice on a 3366, but the same thing still happens. This behavior is far easier to notice on AM010 sensors because the sensor really has no positive quality at all besides hair trigger sensor response being about the closest modern thing you can get to a MLT04 in that regard.

You will be permanently STUCK with inferior cursor control compared to what it shipped with from the factory. Then after you unplug it, it will feel like an entirely different mouse that you will NEVER recreate the fresh out of the box factory movement. Simply take a brand new mouse like a G302 fresh out of the box that's never been exposed to any form of Logitech software (the mouse controls great fresh out of the box), run it through whatever benching/diagnostics, then do the act of just plugging the mouse in to another computer with G-hub already installed (without even changing a single setting, just plugging it in then unplugging it). The results should be easily quantifiable if anyone wants to actually test it (QVC). But this phenomenon with Logitech's new G-hub software is exponentially worse than pre-G-hub software where even a caveman should be able to tell. It's like the method of writing data to the mice these companies are using is completely incapable of any deterministic behavior WHATSOEVER. Even if you're saving a setting of 800 DPI, sometimes it will feel slower and lagging behind, other times faster. So, I have always said that anytime you install mouse software and save settings to the mouse, it always gives you a wildly different result each time whether it's Logitech or some other company's software.

Logitech g hub not starting camera