Thanks for the reply.
That conclusion is simply out of my impression of the coverage of the communicative nature of the Lexical Syllabus in the paper. Maybe I need to read more.
It seems to me that the efforts of the LLC (in the mid-1970s) and Lancaster/IBM SEC are worth considerting as well. Certainly, it depends on how we interpret "较早的、有规模的" , for instance, LLC has only 500,000 words, although being the first electronic corpus to have prosodic features built into it (Garside,etc., 1997:10). If "有规模的" simply refers to the raw size, no corpus can beat the Bank of English. Considering that annotating a corpus is more demanding than assembling raw texts, I cannot completely vote for Birmingham.