Here is an article I read that goes into depth about the problems modern movies have been having with dialogue intelligibility. It's an interesting read with lots of good takes from professional sound engineers.
This really is a problem with many causes, as pointed out. This is part a mixing problem, but equally a speaker problem, I think.
It is easy to make a speaker with clear shouty dialog, but not one that will reproduce natural speech and music.
In the cinemas of old, the speaker system was almost universally Altec Voice of the theater. These speakers had clear dialog, but colored it with a very characteristic Altec flavor. They were definitely "shouty". Music, especially the strings were "stealy".
I think part of the problem is the "Westlake" problem. I have been in a few Westlake studios in my time, and frankly I find them very reminiscent of the Altec sound. So if mixing is done using speakers with large sectoral horns they won't realise they are creating an aberrant mix.
An engineer told me years ago, that you can't help but mix to your speakers. My experience tells me that is absolutely correct. That is one reason why I have been obsessed with creating really accurate speakers.
As far as streaming, it is all over the map. Netflix is really very good. Amazon is variable. I think they have made it so you have to use their app. If I use the Amazon app and eARC the sound is better than from my HTPC. Blu Ray discs are definitely better. We do not watch many movies, but when we do, we order the BD usually.
As I have said before, the most difficult speaker to design is the center. To design a speaker that will reproduce clear, intelligible and NATURAL speech, and also be an excellent music reproducer, blending perfectly with the left and right speakers is a huge challenge. It is for me anyway.
Getting my theater center speaker absolutely correct took a number of iterations. I finally, I think have my in wall center where I want it after two and a half years.
The TL center in the theater though is the reference. Part of that is that the front three are incredibly low coloration speakers, which is in great part due the TL design.
Finally training comes into it. Europe, and especially the UK have renowned colleges to teach the art of sound recording, and not rely on, on the job training.
I think there is a huge difference on the whole from sound mixes in the US and Europe. I would especially site War Horse. This is a noisy movie in parts, but the dialog is clear natural and perfectly balanced. The orchestral sounds are superb also.
This is a difficult problem with major issues at the studio/mixing end and also the home and theater.
