The only benefit of CCD based camera sensors is that they're cheap. SNR is worse, color accuracy is worse, no live view, and slower sensor read-out. There are other drawbacks that aren't due to the sensor tech, but rather due to the age of the rest of the camera. Worse burst rates, slower AF, etc.>>4100265
CCD and CMOS due have very different spectral response without a CFA. This is why most scientific applications have CCDs. https://www.researchgate.net/figure/3-Spectral-sensitivity-graph-of-CCD-sensor-CMOS-sensor-and-human-eye-3-Width-is-5mm_fig3_216533561>>4100533
While color accuracy has gone up with recent CFA design, its not due to less noise, its due to wider response of the filters. Though less noise is benefit of this design choice. https://blog.kasson.com/the-last-word/optimal-cfa-spectral-response/>>4100492
CCD's thick CFA's color response is no closer to slide film than modern CFA response. It was the camera input profiles that gave that look. Even if your goal is crazy color, its better to start with accurate color then go from there. >>4100543
This is not true. Old MF CCDs have massive 9 micron pixels and yet still has worse SNR than any modern baby sensor. Camera processing is much harder with old CCDs due to noise and inaccurate color.>>4103713
Filter correlation with sRGB primaries, or any other working color space doesn't tell you much about color accuracy. This is because camera sensors are linear, and working spaces have gamma. Brightness of colors relative to each(color and brightness contrast) is by far the most important aspect of color accuracy, which is initially determined by the camera input profile.
We also don't want our filters to line up with a working space in the first place, we want our camera to see colors like we do, aka the CIEXYZ color space. This is because working spaces have a defined white point, usually D50. the standard observer accounts for human color perception under any lighting conditions.