I watch YouTube with internal TV speakers and I understand everything, even muddled accents. I cannot understand a single TV show or movie with the same speakers. Something tells me it's about the source material, not the device.
In a movie the characters may be far away (so it needs to sound like that, not like a podcast), running, exhausted, with a plethora of background noises and so on.
If we cant do the same in the movie, sound is just badly mixed. It is not the story setup and it is not "realistic".
Because in real life you don't listen through an internal TV speaker, duh.
That being said, people listening to TV through TV is 100% predictable. If we cant understand, the mix is not "realistic", it is "badly done".
Somehow youtube videos don't have this issue. Go figure /s
The problem, as you say, is that if you don't want to have loud parts, you lower the volume so that loud is not loud anymore, and then the quiet but audible parts become inaudibly quiet.
I consider this to be a separate issue to the lack of clarity of internal speakers, and a bit harder to solve because it stems from the paper thin walls common in the US and other places.
You can usually use audio compression to fix this if you can't play the movie at the volume level it's meant to be played.
Intentionally making audio uncomfortable is not a sign of art or skill, it's a sign of delivering a bad product.
75 dB in real life is your typical restaurant, office, etc.
The most common experience on the poorly mixed content that several in this thread are complaining about are: the volume setting necessary for intelligible audio results in uncomfortably loud audio in other parts.
This is a defect of the content, not of the system it's playing on.
The crux of the issue IMHO is the theatrical mixes. Yes I can tune the TV volume way up and hear the dialogue pretty well. In exchange, any music or sfx is guaranteed to wake the neighbors (I live in a flat, so neighbors are on the other side of the wall/floor/ceiling).