In academic works or projects, 5% plagiarism is permissible; anything above 5% is considered plagiarism and is not accepted
I've never heard of that 5% rule. If you write an article about Bitcoin mining, chances are at least 5% of the words are also used by someone else who write an article about Bitcoin mining. And that's fine.
But if you write 10 pages and copy half a page from someone else, that's not good.
I'm not sure, but I believe the 5% rule applies to all academic journals globally because there's a 100% possibility you'll have to cite or copy a few lines from someone else's work without even realizing it, as you said it happens on a regular basis. No academic publication or paper can pass a plagiarism test with a score of 0%.
Academic work is continued work + added value.