While I understand the survey is designed to be as simple as possible, the details are too simplistic and the instructions need to be expanded or supplemented with FAQ.
For example, the instructions say birds must be 'on the ground'. If participants take this literally, they would exclude birds on feeders and I would not count birds on my garage roof (the only occasional place I see birds in my small garden). Presumably some people will take this literal approach.
In a very large scale survey, small variations in technique don't matter that much, but it is still much better to be clear.
Another FAQ could address where male and female birds are clearly different. If i see first a male blackbird, then a female, i count 2 blackbirds. Other people may count 1 as the maximum.
Then there are some people I know who will count whatever they can see from their window and others who are rigid about only counting the birds 'in my garden' as opposed to on the fence or in next door's bush.