I'd like to comment on your points one by one, and I'm going to that here for it is here that they have the most context. If you'd rather not receive comments on this page, I appologize, and feel free to delete this post, or transpose it to your talk page, the verifiability talk page, or nowhere:
I've had to explain this exact point to people on multiple occasions. I feel the emphasis on verifiability has more benefits (by discouraging newbies from trying to remove "falsehoods" from an article when they've actually removed commentary on notable falsehoods or more ambiguous disagreements, and similar problems) than drawbacks (such as some editors treating the policy as a suicide pact, that we must retain false information even if we can objectively out an error in a source). Perhaps this could be elaborate, but I have always just ignored as necessary.
The strict adherence some people have to this (as well as the strict ignorance of it) is the cause of many edit wars, but I see no consistent way out of it aside from dispute resolution.
Mainstream newspapers have never seemed problematic to me, as I have witnessed many others profess. There is a common error in judgement that newspapers are incapable of providing accurate information on academic subjects. While certainly new sources are prone to oversimplification and sometimes inaccurate explanations of academic topics, any good news source is very clear on what it is claiming as a fact, and what it is claiming as a find or an opinion. When a news source reports an opinion, many editors are unfortunate to treat the opinion as a fact, and others are equally unfortunately to claim the news source unreliable for reporting a "fringe" idea. While this may not be what you were thinking of at all, I've noticed these as the most frequent problematic interpretations of that line. As to the bias of a news source, I've entertained the possibility that news sources alone do not establish the significance of an academic opinion where contradictory opinions have a clearer significance.
I have no suggestions on "respected," and this is one that has also forced me to reach for a definition on an as needed basis.
It's more ambiguous than it is incorrect, but I'd also presume it's an accident.
The expert clause is aimed at academics, and not television writers. This is a failure of the letter to be nearly as confined as the spirit.
I believe the main hurdle to overcome in bypassing that bias is ensuring the policy doesn't become too loose. There is usually little difficulty in branding something a reliable source/self published source/expert source, or at least it is usually easy to reach consensus. In granting greater policy-abiding latitude to use what were previously considered unreliable sources, we don't want to fall into a hole wherein every post a certain individual makes on his blog, or every word out of his mouth suddenly becomes reliable and citable material. I wish a case by case basis would always work, but I'll admit to seeing an unfortunate number of editors blindly obey the policy without truly considering the possibility it could be ignored.
The examples you suggest remind me of press releases. Certainly, I would hope no one considers a press release from an actor's agent an unreliable source because it's about someone other than its literal author. You'll also notice there have been a number of attempts at rewording this section in just the last several weeks. I can't decide yet a great way this could be worked in. Perhaps the ban could be lifted for publishers "intimately associated with the living subject," and of course a warning not to use this to sidestep BLP. Anyway, the point is that we can trust it to be an accurate primary source on the person.
