They were living in the age of circulating pennies, and these figures represent absolute reality at the time. There's no way such a feel could be reproduced now, post decimalisation and post melt, over 50 years on. That's why these surveys, however rudimentary some of them were, are so historically valuable IMO. Actually there's nothing wrong with rudimentary - it's tabular. straightforward and very easy to understand the meaning in a few seconds, which is what you want. I wish more stuff was like that these days.
Yes, he's obviously got the estimates for 1919 & 1919H the wrong way round. Also, as you say, quite why he would have nil as the estimates for 1922 and 1926 is a bit odd. What he actually found was (I reckon) pretty much in line with the sort of figures out of 10k, you might have expected for those years. Possibly even a tad under.