Isn't there some advanced software that academic institutions use to detect plagiarism? Wonder if we can get an overview of the algorithm used by that software. That could easily govern the set of rules that should form plagiarism.
Even if such a software exists, it should use the famous "five consecutive words" rule. Also, it could be some set of plagiarism rules implemented by the software developers. I think the plagiarism rules frequently adopted are not that universal based, rather based on ethical considerations. I think some plagiarism rules in this forum are not cast on stones but rallies around the intention and level of prior positive contributions.
If some strictness would be applied in explanation rather than implementation it will be fine cool.