I read the article you linked; what a herculean (aristotelian?) taxonomical task. Probably could not have been automated.
I just found a fabulous article by Christian Kay which summarizes a lot of interesting things about the project, including the paper/technology issues - - much to interest readers of this list:
<http://eprints.gla.ac.uk/49097/1/id49097.pdf>http://eprints.gla.ac.uk/49097/1/id49097.pdf
I know from talking with the designer of the book that the design process alone evolved over more than 20 years. One figure that Kay reveals is the cost of the project: about a million pounds, or about a penny per entry, spread over 44 years. Here's another gem: "There then began a short break while the volumes were typeset abroad, although any relief this provided was marred by a cautious member of the team calculating that the typesetter's claim of a '99.995% accuracy rate', when applied to HTOED's 22.74 million pieces of data, would result in over 1,100 new errors being created."
I agree with your assessment that it was a herculean taxonomical task. Kay is very modest indeed about her life's work, but I can't help feeling that she should be ranked at the same level as Murray. Why is it taking so long?
Partly it's because the work is difficult to figure out at first. I got my wife to request the review because I wanted the book. But once we got it, we had no idea what to do with it, and there were no other reviews to help. The promotional material is unhelpful. Gradually, we began to understand, but it did take a few weeks; not just to get it but to get comfortable with it.
