My perspective on color CMOS cameras is they have their own Genre like Spaghetti Westerns. (Good, Bad and Ugly fan) Trying to replicate narrow band images like a mono sensor with filters is problematic at best. As far a color saturation is concerned with broad band RGB images, I'm not sure about that. I do know that imaging with them took a rethink of my normal CCD imaging techniques. They gather data so fast, it's easy to start decreasing your signal to noise by exposing subs too long. In that case, long sub times may actually kill the image in brighter skies. And of course, there is the band passes of the Bayer Matrix filters to consider vs individual traditional RGB filters for mono camera.
You need to break up your integration time into a larger number of subs of shorter duration if you are broadband imaging. With narrow band OSC filters like the LeXtreme Pro (7nm band pass on Ha, OIII) it doesn't matter very much. Of course you need long integration times, but at the proper sub time. This of course is problematic with smaller pixel Cameras because the files are so large. But computers and SSD drivers are cheap, and new sensors like the IMX410 with 5.94U pixels give me hope that larger pixel CMOS sensors may eventually appear because they are optimum for commercial low light camera applications too.
I've been using a portable SQM meter with separate Red, Green, and Blue filters over the lens (simulating what the OSC sees) while stacking various OSC filters over them to take sky SQM readings. I then use that to determine optimal sub exposure time limits for my CMOS OSC. Of course there is a lot of wiggle room there too. It's almost impossible for me to image with CMOS color cameras in Bortle 8 without some kind of light pollution filter. Sky fog can be an issue after 5 seconds easily causing washed out images. It's really amazing how much light pollution filters reduce the SQM readings of the bright city sky with great transmission. With my sky MPSAS of 17.87 (Bortle 8.1), the sensor under the red filter of the OSC matrix and an IDAS D2 filter sees 20.52 MPSAS or Bortle 4.10. Of course they filter out good colors too.