Also:

If one treats manuscripts as an epigrapher would treat inscriptions,
facsimile editions of manuscripts as they were actually written
can provide very useful information:

1. Changes in writing style can be used to date manuscripts.
2. Colophons that provide authorship information
and context for the work have usually been left out
of edited works because they are not considered part of
the one unique universal version of the text.

e.g. if one used PTS transliteration for the Gandhara texts, for instance:
http://en.wikipedia.org/wiki/Gandharan_Buddhist_texts

One would lose valuable information. Most palm leaf texts
are fairly late copies but some date from as early
as the 1400s and caves have been found with manuscripts.

Extensive inventorying of manuscripts throughout Asia has been done by
Filliozat and Skilling who are both members of the French EFEO.

EFEO publications always include versions of the text in localised
scripts. For example, Bizot and Lagrirarde's translation of the
Saddavimala in La Purete par les mots (1996). Peltier does also.

The study of manuscripts and scripts is still in its infancy in
Buddhist Southeast Asia. Southeast Asian scholars argue whether
Burmese script came from Mon or vice versa, but that they both came
from a Pali script used in a different context seems more likely.
Given that palm leaf manuscripts are very perishable, only the script
of inscriptions or votive tablets will likely provide proof.

With metta,
Jon