In Cornell's paper, could all seven of the sight records really be mis-IDs?
I think the answer is "Yes".
Remember, Cornell described them as "Fleeting Glimpses". These were typically brief, flying views, at a distance and/or with the naked eye.
As Gallagher relates in "The Grail Bird", page 164: "If you want to see an ivory-bill bad enough, a crow flying past with sunlight flashing on its wings can look pretty good.
I think an honest question-and-answer like this could follow most/all of the reported sightings.
1. Did you see the underwing well enough to describe it correctly?
2. Did you see the white dorsal stripes well enough to describe them correctly?
3. Did you see the bird well enough to describe the white neck line that ends before the bill?
4. Did you see the bird well enough to describe the pale bill?
In my opinion, if you answer "No" to all four questions, you didn't see an Ivory-bill really well. I can't imagine a state records committee that would approve such a sighting.
In the book Sibley's Birding Basics, a couple of key paragraphs on page 41 may be relevant:
...An observer might see something intriguing, say a large falcon flying away, and jump to the excited conclusion that it could be a Gyrfalcon, a bird normally found in the far north. The next step should be to pause and start from the beginning, looking at each characteristic objectively, but too often the overexcited birder tends to stick with the first impression and simply tries to confirm the identification as a rare species. Often, a very brief sighting does not allow any more detailed study, and the observer might choose to emphasize anything that can tip the balance toward the desired identification: "Yes, it did look long-tailed; yes, it was very dark; it just didn't 'feel' like a Peregrine,", and so on. Other poorly seen field marks that point toward a Peregrine Falcon--perhaps it looked long-winged, or seemed to have a contrasting white cheek--are then ignored.
This problem can result in a sort of "group hysteria" when large numbers of birders look at the same bird. The suggestion by one person that the bird is a certain species forms an expectation for everyone else, who then looks only for field marks to confirm the "expected" identification. In one very well documented case in California, the first state record of the Sky Lark (a Eurasian species) was misidentified for days, and by hundreds of people, as the state's first Smith's Longspur. The two species have a superficial similarity but are not even in the same family and can be distinguished by dozens of features. The initial observers expected a Smith's Longspur to show up in the state and never considered the Sky Lark as a possibility. Most of the people who went to see this bird over the next few days had the same expectation, augmented by the knowledge that they were looking for a "confirmed" Smith's Longspur.
From quotes I've read, in the spring of '04, it seems like conditions were ripe for mis-IDs as discussed in the Sibley paragraphs above.
Ron Rohrbaugh said, "It was an absolutely electric time. To think that around every bend, behind every big cypress, there could be an ivory-bill."
I'm not up on a pedestal saying that it couldn't happen to me. A painfully embarrassing case occurred when I was glassing for wildlife from a high hill in Alaska one fall. With a previous sighting of distant black wolves under my belt, I again saw some slow-moving black dots in the far distance. After I shouted to my companions that I was seeing wolves, we discovered that the black dots were actually ravens.
It happens, even to the "experts". If you combine a brief, distant glimpse of a flying bird with the expectation that the bird may be an already-confirmed rarity, you have a potent recipe for a mis-ID.