This is looking very polished.  Thanks to Adam and to the mailing list
participants for this clear and detailed SRFI.

A plug-and-chug test run on chibi-scheme resulted in all tests passing
without issue.

There are a few minor typos in the SRFI document:

* The spec of flexvector-cumulate reads:

  "Returns a newly-allocated flexvector new with the same length as
  fv.  Each element i of new is set to the result of
  (f (flexvector-ref new (- i 1)) (flexvector-ref vec i)) ..."

  For ‘vec’ in the Scheme snippet, read ‘fv’.

* For clarity, flexvector-index should indicate that it applies
  ‘pred?’ in parallel to the elements of the flexvector arguments.
  (The specs of most similar procedures in SRFI 214 make this clear.)

* string->flexvector's ‘start’ argument should default to 0, not -1.

Minor quibble: Why not provide the inverse of generator->flexvector?

Best regards,

--
Wolfgang Corcoran-Mathe  <xxxxxx@sigwinch.xyz>

"All this currying's just a phase, though it seldom hinders."
--Fritz Ruehr