A straightforward exercise and something I’m surprised I’ve never actually done before. For it I used my Nikon D50 body, Nikkor AF-D 35-70mm f2.8 zoom at both ends of its range and a Zenitar 16mm f2.8 to give a final, wider version of the view. I was interested to see that – when looking through the viewfinder with the other eye open – sizes didn’t match at a focal length of 35mm (the normal/50mm equivalent for my camera’s sensor) but did when the lens was zoomed in to about 50mm. While the angle of view was correct for the 35/50 equivalency, optically the perceived size was correct only with a tighter focal length. Presumably this means that my “sensor size” (the retina of my eye) is closer to the frame size of 35mm film and while it is seeing a narrower angle through the viewfinder (approximating to a 75mm lens on a film camera) the visible size of the objects seen remains the same. Certainly, a couple of days later, trying a 50mm lens on a film SLR, it was still the 50mm lens that gave the same perceived size in both eyes. Now, it would be worth trying this with a medium format SLR and a 50mm lens, just to check whether a 50mm lens – acting as a wide-angle this time – would show objects the same size in the viewfinder as they appear through the other eye in reality.
Anyway, here is a composite of the three shots I took, with the tighter angled pictures scaled down to the equivalent size in the widest and then pasted over the original with the frame’s edges outlined:

When I returned with 3 A4 printouts, I was just able to get my hands far enough away from my face for the 70mm picture to match the size of the original, the 35mm picture was ‘right’ with a slightly obtuse angling of my elbows and the 16mm picture would have needed to be too close to my face for me to focus.
Of course, I forgot to take a tape measure with me, but – back home with a ruler and relying on muscle memory, or whatever it’s called – I was able to estimate the distance at which the prints matched up in size to the view as being 70cm for the tightest and 35cm for the middle one. The widest was impossible to gauge, but it seemed to be almost right when the paper was around 10cm from my face; from the other two, it presumably would have been about 16cm given that (for an A4 print) focal length in mm equals arm’s length in cm…