These contexts don't change the idea in rule 1, but they do change the definition of what is "necessary". There are words that "dilute" a message and are not necessary to convey the literal meaning. But usually we don't want just the literal meaning of things, we want a certain level of dilution. Without it, prose can be clear and specific, but at the same time becomes dense and unreadable. Lots of academic writing is like that.

There are words that are necessary for the _overall effect_, even if not necessary for the literal communication goal of a set of words.

To be clear, this is just a simple summary of the source. I agree with your statement, though. In fact, as Rudolf Flesch said in his book, How To Write Speak And Think More Effectively, "So here we have the secret of plain conversational talk: it is not difficult ideas expressed in easy language, it is rather abstractions embedded in small talk. It is heavy stuff packed with excelsior. If you want to be better understood you don’t have to leave out or change your important ideas; you just use more excelsior. It’s as simple as that." The point his is making that the few clear simple points can be packed with some excelsior to improve clarity.

There will be exceptions to all these rules at the expert level, but less expert people should probably follow them.

I was reminded of this recently after reading an essay by a high schooler. The essay was actually pretty insightful, but I couldn't help but notice the way in which the words that the author said seemed to be a great deal more than strictly necessary to communicate the point that he was trying to make to the audience that he was speaking to at the time he was writing the essay.

Perchance.