I'm not really sure what Pugh's getting at here. My best guess is that he's worried that one might have trouble distinguishing
"For all $x$ there is some $y$ such that [stuff]"
from
"There is some $y$ such that for all $x$ [stuff]"
on the grounds that "for all" sort of sounds like it's describing a single condition which must be met all at once.
OK this doesn't quite match up with Pugh's "more inclusive connotation," but it really is the best guess I have.
But to me this is a huge reach: quantifier alternation is a genuinely difficult topic for many people, language be darned, and I don't expect "For each" to do much better than "For all" in this respect. Certainly nobody will look at you askance if you read "$\forall$" as "for all."
Conversely, nobody will look at you askance if you read it as "for each" either. If that choice of language makes quantifier structure easier to parse, you should absolutely do it. But there's nothing wrong with "for all" here.