I am increasingly testy about the modern habit of printing everything very large. It strikes me that this is an effort to wow the viewer with a mass of detail and the sheer bigness of the thing, and to conceal thereby the lack of content.
There's more to it, though. There is a definite field of view that we can apprehend "all at once" and it is surprisingly small. By making large prints and then hanging them in such a way as to force us quite close to them, the artist alters the way we see the work. We have to scan it, and mentally assemble the image. Our visual cortex helpfully does all the heavy lifting for us, and in general we're unaware of the scanning process unless the print is very very large indeed.
Let's imagine we're in a museum, looking at prints on the wall. Let's say we're something like 8 feet away from the wall. My experience with museums suggests this is a pretty typical viewing distance, plus or minus a few feet.
The width of the visual field we're conscious of is something like 40 or 50 degrees wide (this is the view a "normal lens" gives on a camera). To see a wider area than this, we deliberately turn our head from side to side. In the museum, if the print is more than about 5 feet wide or so, we're going to have to start doing this, or we need to step back.
The field of view we can actually grasp "all at once" in some useful sense, seems to be much smaller. Let's say about 5 degrees (experiment yourself, fix your eyes on the page, estimate what you can actually see clearly without letting the eyeball move, and do the math). This corresponds to a print size of about 8 inches across at our 8 foot viewing distance. Call it an 8x10 print, since the stuff around the edge of the frame probably isn't very important. We can grasp the essentials, the central portion, of an 8x10 in one go if we're 8 feet away. More or less.
There's a great deal of daylight between the 8x10 print and the 5 foot-wide prints you have to consciously scan to grasp, and this is pretty much where everyone is printing these days. With the typical 16x24 and up prints, our visual cortex is working harder. We're assembling a notion of the forms in the print with peripheral vision, and unconsciously scanning to fill in detail here and there, and over the course of a couple of seconds building up a clear picture of the image. What's really interesting here, by the by, is that it feels instantaneous because we're editing our memory to obscure the time lag. Look up chronostatis for the details.
Finally on to the subject of this piece. Thumbnails are a new thing. They belong to the Internet, they originated there and are a uniquely digital artifact of the way we organize and present images. They're small, which is the point. They're small enough to apprehend in one go, we see them in an instant. While we see them in an instant, they also have certain properties akin to a very large print, in that they are really a low-resolution summary of the image in the same way our peripheral vision provides a low-resolution summary of our current field of view.
It doesn't happen to me very often, but occasionally the thumbnail presents a very clear idea of the image I am about to look at, which is completely wrong.
Always, the thumbnail provides a wildly different experience of an image. In pretty much all cases, I find that the thumbnail tells me if the photograph is any good. That is, if I like the thumbnail, I will like the photo, and if not, I won't. But not always.