This is an old argument and involves srfi-133 as well, but I didn't recall it
before I stumbled on it again.

@vector-fold's kons takes seed argument first, followed by the element.
It is reversed from srfi-1 fold and srfi-152 string-fold.

gosh> (fold cons '() '(1 2 3))
(3 2 1)
gosh> (string-fold cons '() "123")
(#\3 #\2 #\1)
gosh> (u8vector-fold cons '() '#u8(1 2 3))
(((() . 1) . 2) . 3)

@vector-fold just followed srfi-133 vector-fold.  I wonder why it was so and dug the
archive.  It seems that the issue was raised but hasn't been resolved:
https://srfi-email.schemers.org/srfi-133/msg/4224107

The order of kons in vector-fold / @vector-fold is the same as R6RS fold-left.
In a retrospect, vector-fold-left and @vector-fold-left would've been a better choice.

Now that srfi-133 is (scheme vector), so it would be difficult to change.  It isn't desirable
that srfi-160 and srfi-133 deviate.  I'm not sure how to solve this.   At least, if there
would be another chance, I like it to be adjusted for the consistency.