Because camera data is already encoded logarithmically in common image and video formats.
The article didn't explain what its "log" is at all. Is it the same as gamma? The same as HDR? Something professional? Or something new?
If you don't know anything about photography, it seems like they're explaining something new. If you do know things about photography, you realize the article is full of buzzwords that aren't actually explaining anything at all.
The main benefit of log as opposed to gamma encoding I can see is that log shares the nice property that each f-stop gets allocated the same number of bits; this results in the fact that while gamma encoding tends to allocate less bits to the brightest parts of an image to allocate things in a "perceptually" fair way (at least according to Steven's power law model), log encoding tries to allocate things in a "physically/optically" fair way, which I guess ends up working out better for editing?
Of course in terms of benefits I suspect that "log" mode also reduces a lot of post-processing.