He has great numbers, up until the end of 1917 where it was 2 900 000.
It says 5 000 000 till the end of the war, I think the source is a from a 1920s reference.
I really, really dont believe the 5 000 000. The bookkeeping up until the last quarter of 1917 was really good, then it seems to be a wild guess and the proportions just dont work out.
It is a really, really great book, but i have an issue with the 5 000 000.
(My issue is that using about 750 award dates, I calculate proportionally how many were before and after the last accurate number in late 1917... and it is less than the 5 000 000)