If a scholarly commuity cannot identify what material it studies, it has not yet matured to a point where digital technology matters very much: scholarly discussion requires being able to cite evidence. I worked on the CITE architecture for scholarly reference in part because I come from a background in Classics where, by and large, we have a reasonable tradition of citing works by logical, canonical reference schemes. (Of course there are exceptions, like that little corpus of Plato that we continue to cite, bizarrely, by physical pages in the sixteenth century edition of Stephanus...)
The suggestion that scholars need to be able to identify and cite their evidence seems to me a pretty minimal measure of the maturity of a discipline, but if I hold up classicists' conventions as a positive example of canonical citation conventions, people occasionally misunderstand this as an elitist attack on their field of study. No: I want to apply the same standard to subjects I work on.
For example, in the study of ancient science, we have not advanced much beyond the stituation described almost 40 years ago by Neugebauer (never one to sugar coat his judgment of the state of scholarship) as follows:
For classical antiquity and the Middle Ages no systematic collection of mathematical or astronomical treatises exists. No attempt has ever been made to compile basic collections comparable to the Loeb Classical Library or the Budé Collection, or Migne's Patrologia, the Monumenta Germaniae Historica, the Bonn Corpus of Byzantine historians, etc. ... This fact alone suffices to show that the so-called 'History of Science' is still operating on an exceedingly primitive level.
A History of Ancient Mathematical Astronomy, 1975, p 15
Can we leap straight to a digital corpus for ancient science, like a third-world country bypassing costly and slow expansion of landlines and immediately delivering phone service through cellular networks?