No, but it's the job of a standard to specify how to talk to the common
case of a vaguely POSIXy filesystem (whether that standard is part of
the language or something like an SRFI is a matter of definitions, as
long as Scheme apps that want to interact with the filesystem can do so
portably, where such a filesystem exists)
Doing half-way jobs half-way only produces a lose.
I would be completely in favor of a careful Scheme binding specification for Posix. That would be wonderful. It would not involve re-writing Posix, even where Posix gets things wrong, and it would not involve trying to vaguely map to vaguely Posix but not actually Posix systems. If this is to be useful, it should support all of the features that Posix supports, since the point is standardization not design.
The level of separation between what ANSI C specifies and what the C binding for Posix.1 specifies is extremely good. I would strongly encourage taking seriously that they may have distinguished well between what is a "language" issue there and what is an "operating systems" issue. For example, ANSI C has some file operations, with stdio being able to open files in various ways and operate on them, but does not have filesystem operations like directories and such.
Thomas