I've always been dubious of counts that involve a stream of passing birds, largely through personal experience. I remember once being at Portland with two friends. They were seawatching from the obs balcony one evening. Manx Shearwaters were passing by at a reasonable pace - all close inshore and all heading left-to-right. I was jotting down numbers as they called them out every five or ten minutes. "8... 30... 15... 17... 9... 32... 16... 16... 8... 27... 15... 15... 9... 31..." and so on. There was clearly a pattern emerging. I grabbed one of the scopes, pointed it a few degrees higher and told my friends to watch what was happening further out to sea. There were Manx Shearwaters, all heading right-to-left; first a flock of about 8 birds, then a flock of about 30, then about 15, then about 16. The same four flocks were flying round in a big anticlockwise circle. Had we not looked further out to sea, the total count could have been nearing four-figures by dusk!
The highlight [at Spurn on 28th June] was 3 ALPINE SWIFT sightings during the day, and it asks the question is it just the same bird going round? Well we can’t prove anything but we think there probably were three different birds today as there was a very heavy passage of Common Swifts moving south during the day. 8,000 flew south, flying low, direct and straight towards Lincolnshire, with no birds appearing to be u-turning.
I've heard of this thing happening to other people too: a partially leucistic bird seen several times over a morning's seawatch, despite the fact the it was always seen flying one way and never seen to fly back the other way; and other similar stories.
So what of the Spurn swifts? It strikes me that the '8,000 Common Swift' argument might be slightly flawed; what if there were only 2,600 (8,000 ÷ 3) Common Swifts and they all went round in a big loop three times, each time bringing the same Alpine Swift with them?
I know I'm saying 'only' like 2,600 swifts is an insignificant number but the principle should stand that dividing the total number of swifts by three acts to divide the probability of three Alpine Swifts by three. And a big loop with birds flying through Grimsby and back over the Humber further up wouldn't require individuals to U-turn for them to be double (triple) counted. That's my rather sceptical take on it, anyway; I'm more than happy to receive angry comments from East Yorkshire locals pointing out why my theory is completely and utterly wrong!
And as for vis'migging and the difference between a bird that's actively migrating and one that just happens to be flying over... Don't get me started on that topic!