Cel animation had that problem. Like, if there was a scene with people trapped in a bricked-up room, you knew which brick was the door to a secret passage because it was a slightly different color than the rest of the bricks. This was because there had to be two cels, one for the background and one for the moving brick. Even if both cels were drawn using the same color value, it affected how the light cast through. Same thing with tinted windows, if you wear sunglasses too it'll be that much darker.
CGI is created independently of the actual film stock, so coloring is recreated then overlaid onto of the film image. Sometimes the CGI house comes really close, or even nails it, but the human eye and brain are so good at noticing disparity that for some people this difference really stands out. Even if it's just one little thing -- lighting, color, motion, focus, clarity, variation -- the illusion is ruined. This problem has been reduced significantly with the advent of digital film but fidelity still depends on the artistry. The question becomes do we recreate the image the way the eye would see it, versus the way the camera would capture it? What's the priority? Storytelling? Looking cool? There's a lot to control and thus a lot to get wrong. With fast approaching deadlines, there's never enough time to get it perfect.
Even Michael Bay continues to finetune his films after they've hit the theater, for the home market.
Plus special effects guys are not working with a movie theater sized monitor. Equipment affects. Not only in the making of the film but also in the broadcast. Most heavy CGI films don't look as great on a television broadcast because the output just isn't strong enough to maintain the contrast you're looking for. Even AVATAR and MAD MAX FURY ROAD suffer from wash-out.
Smoke and haze is difficult because generally speaking you can't see through smoke. Even thin smoke, the lens has trouble piercing through it. The camera captures no visual information. What should happen is the image is obscured in places and not so much in others, but CGI tends to add smoke that's just this blanket screen with a lowered opacity. There's no waver. Too much clarity is often an issue with CGI movement, which was quite problematic with the MATRIX fight scenes where the multiple Smiths' movement looked clunky. There are people who move so quickly the camera can't capture it, so when you see every single pixel in startling clarity, your brain's gonna go WHOA WTF.
Even if you may not know it consciously, your brain can still tell when things are "off" or too perfect.