What he did do was amass an extraordinary corpus of scholarship on the subject which has not been equalled to this day. Categorisation is a matter that continues to perplex, and opinions are still divided; his system was by no means arbitrary, but can seem impenetrable, particularly as he died before he was able to complete his Introduction to the work, which would presumably have detailed his priorities and general approach. The only other attempt at a classification system, that of G. Malcolm Laws, is if anything harder to understand. I don't yet know what approach Steve Roud is taking to classification in his ongoing Index, so I can't comment on that.
There are certainly a (relatively small) number of ballads (as opposed to folksongs) which are generally considered to warrant a place in the corpus; The Fowler/ Molly Bawn/ Shooting of His Dear and Bruton Town/ Bramble Briar (etc) being two, but Child did pretty well in the circumstances. If he had also been out there collecting material directly from tradition, he wouldn't have had time to compile the books; he encouraged others to collect, however. He died before the collecting boom of the early 20th century happened, of course.
There is probably no student of English-language folksong, and few practitioners of it, who are not indebted to him. Such a work can by its nature never be complete, and his sudden death didn't help; obviously it would be a mistake to imagine that ballad scholarship ended with him, and many who have come after him have expanded our knowledge and understanding beyond what he was able to accomplish. The fact remains, however, that they are standing on the shoulders of a giant.