I don't know when Mortensen began to recommend
underexposing but films even as early as the 1940s had
pretty good density range, well beyond what can be printed
on a reflection print. I have no idea what he thought he was
doing.
  The characteristics of at least Kodak film was published
back to the 1930's. Kodak Labs did extensive research into
sensitometry as basic science needed in the production of
film and paper. Probably the best known papers are by Lloyd
A. Jones and various associates. Jones began research into
tone rendition in the 1920s and devised the system of speed
measurement used internally by Kodak from about the mid
1930's. This system was adopted, with some changes, as the
first ASA speed system in 1943. Jones was interested in how
film was used in practice and his speed system was based on
obtaining good shadow detail rather than finding the
theoretical maximum speed, which was what most of the
earlier systems did.
  In a landmark paper on tone reproduction published in the
mid 1940's Jones detailed an extensive project to discover
what the minimum exposure for "excellent" prints was. He
decided that setting the deepest shadow that was to have any
detail should be on the toe where the contrast was not less
than 1/3rd that of the straight line portion. He further
found that increased exposure, up to several stops, made no
difference in the tone rendition of the print, once the
minimum was passed.
  Unfortunately, the ASA, perhaps at Kodak's urging,
included a 2X safety factor in the speeds resulting in
overly dense negatives. When the second ASA sytem was
adopted in 1958 this safety factor was dropped so all film
speeds doubled overnight!
  My point here is that there is substantial evidence that
films of the 1940's had plenty of overexposure latitude and
not much problem from running into the shoulder in any sort
of normal pictorial photography.
  It is quite possible that much earlier films did have
such problems. The improvement of film throught the 1930's
and 1940's was tremendous. One evidence of this is motion
pictures. Assuming you can see and hear a print that is
actually representitive of the time, you will find that both
photographic quality and sound quality improved very quickly
from about 1930 to about 1940. This was partly due to
several discoveries about sensitization, allowing finer
grain and higher speed, but there were other factors as
well, for instance gelatin formulation and new ways of
hardening the emulsion.
  In any case, I stand by my judgement that the method of
"gamma infinity" development Mortensen preached probably did
not work very well at the time and certainly is of no more
than historical interest now.
  I don't know if any of Mortensen's books have been
reprinted. If so, they are worth reading there really is
some good stuff in them. The original editions have become
rare and quite expensive. Some libraries may have them
available. However, the man was mostly a charlatan, albeit a
talented one.
  Some details of Mortensen's "Abrasion Tone" print
finishing process can be found at the Bostick & Sullivan
site at: http://bostick-sullivan.com
  Another process he used often is the paper negative
process. This is not hard to do and RC paper works fine for
both the paper positives and negatives. It produces
interesting prints and can be used for very extensive
retouching and modification of prints.