With the appropriate lens, the number of available pixels or sample points for recording (Figure 3A vs 3B) dictate how much detail can be captured. Figure 3A & 3B, both use a 4bit gradient, meaning both can can show 16 shades of grey. But 3B has more pixels compared to 3A, so it was able to show more detail because having more boxes to fill, it was able to use of more of the available shades.

On the other hand, the software (together with the sensor/electronics technology & pixel size) will indicate how much Color depth/gamut is possible (gradation/shade recorded per Bit - Figure 1,2 & 3)

The computer records them as groups called bits, like the 8bit/12bit/16bit/jpeg or RAW files where each bit is a group of bytes corresponding to a single shade. Higher bits produces a smoother gradient transition.

Figures 1A, 2A, 3A & 3B Shows how each would show a black pearl image.

Please note that this is a very simplified explanation.
Sensor only capture intensity of light. Hence, the matrix of the sensor are divided where some pixels have filters that allow only the green to pass, to output green signals. Others have red and the remaining with blue. Each one giving its share of the RGB signals whose blending produces the other colors/shade.

Technically, a 1megapixel camera, is only a 0.33megapixel camera. Because only 1/3 of its sensor record records each color*(not exactly true because some sensors have filters that favor recording more of one of the colors, such as the Bayer filter which allot more pixels for recording green). A group of 3 colorpixels are needed to show each point of color in the gamut.

To combat these inefficiencies, various software filters and algorithms like smoothing, dithering etc. are applied to the output. New technologies are also being explored such as pixel shifting & colored sensor layering

Yea, just like megs and gigs....

We (programmers) said 30K for a file that was 30 Kilobytes in size, NEVER Ks.
It was a 20Meg file, NO "s". (Meg stands for megabits, which is already plural.)
The media and people who were NOT in computing started with megs and gigs as plural. Guess what became the "standard"? I STILL say 25Meg and 5Gig.
I can't wait until people start using ters for terabits. Why haven't they? Sound stupid?
(Sorry, been a hitch in my git-along since it started.)

Unless they dropped it, a "word" could be either 4, 8, 16,... bits, depending on the architecture of the processor. Many micro-controllers started with a 4 bit word. The Z80, 8080, 6800, etc. were 8 bit word processors. Then came the 16 bit word processors.