John Mark Ockerbloom’s conclusion that
In libraries, size of your collection is important, but even more important is what your readers can do with the collection. In the early going of mass digitization, quantity makes the big headlines; in the long run, improving quality may well have the greatest impact
leads me to wonder what tools could we integrate in PReE to improve the quality of a corpus? Should we attempt to pipe improvements to other services or repositories? How can we know what is an improvement versus what is an error?