I do know the origin, but if we permit any influence of descriptive linguistics at all, we have to admit that people are not substituting et for and and then re

I do know the origin, but if we permit any influence of descriptive linguistics at all, we have to admit that people are not substituting et for and and then rendering et as &. They are directly substituting & for and, so it’s a distinct usage imo. Yes, the historical reason they make that direct substitution is because they have read enough writing from authors making the indirect, via-Latin substitution to deduce the apparent meaning of the symbol. You can also support the claim that the direct substitution is distinct with some experimental linguistics: find someone using & and replace it with et, they will tell you that’s wrong, but replace it with and and they’ll be okay with it.