[anikom15]

It’s because the viewing format has a fairly consistent gamma curve relative to viewing environment ambient light level. In a theater its about 2.6 and this works with no ambient light, and in a living room its about 2.4 with dim lighting, and on a computer screen its about 2.2 with typical office lighting. With controlled lighting, the perception of the output image under these three different gammas will be identical. This gives graders a great deal of precision. They can calibrate their screen and studio to appear exactly as it should to these idealized environments. That a number of people watch content in non-ideal environments is irrelevant (even though its often brought up as a strawman argument against consistent working environment viewing conditions).

With photography, you don’t have that control at all because traditionally photography was about prints. Every printer (regardless of kind) had to be profiled and you needed an explicit ICC conversion process for it to work. If you are solely publishing photographs digitally and have no intention of printing, you actually could adopt a film workflow and it would be just fine.

In short, film production doesn’t actually need the added complexity of ICC profiling. They could adopt it, but it’s not strictly necessary. Trouble comes when people don’t understand the fundamental reason why gamma correction and different primary spaces exist in the first place.