Our and-let* wasn't entirely conformant to SRFI-2.  It would return #t
for a variety of forms where it should return the last evaluated
expression's value instead.

Here's a patch.  Is the copyright a problem?