Bokeh highlights project the light spots as the shape of the iris, in the human case, pupil. This is round and stays that way when you stop down. A cat would see them as oblong pointy ellipsoids. The next feature that decides the bokeh look at spherical aberrations, which are more difficult to predict. Maybe this component is affected if you use glasses, especially if you need correctional glasses like I have 60degrees on one eye. I found this text:

http://www.telescope-optics.net/eye_aberrations.htm

Apparently we normally do have nice correction and then it should fade towards the edges, like the best of the best lenses. And when you are near and long sighted yours focal length and eye ball distance dont match.

Another study finds a bias towards negative spherical aberration, which makes better defocus (bokeh balls) edges when focusing closer than the object.

I've done some tests and agree that the bokeh ball is near perfect round, but has a texture on it. This must be due to the uneven distribution of receptors, organic matter, and the "blind spot" we have where the nerve connector is.

The third thing that affects the view of the balls is the F number - relative size of the iris to the focal length. Research found that pupils are around 4-9mm depending on age. The focal length (not 35mm EQ!) is 17-22mm depending on how you measure it. This gives a span of F1.8-F5.5, so in most cases we would go around with F2.8-F4.

http://hypertextbook.com/facts/2002/JuliaKhutoretskaya.shtml

The balls created by focusing closer then the highlight are rather small and fading to the outside edges, making them seem even smaller than the balls e.g. here at F1.4:

But a laser pointer at 3 meters distance did become a handsized bokeh ball when I focussed at 30cm. My eye that has 60deg correction was not making a uniform ball while my other eye did.

The Focal length is also in the equation:

http://www.marcuswinter.de/archives/1703

When using the F number as reference (your F3.5-5.6 zoom is really a fixed aperture lens, and your F2.8 zoom is variable aperture, but fixed F-number), it scales linearly with focal length (mult. crop factor) and the human vision is considered close to 50mm (eq.) , which the dog picture was taken at, so for the sake of simplicity lets not account for that:

So we would see the balls in the dog picture about 20-50% the size, a bit rounder, fading to the edges, and with textures inside.But if you have the highlights closer then your focus, the balls take a hard-edged look and seem larger (probably also due to closeness).

This is all theoretical from my part, so take it for what it is. I might investigate further and update with my findings (again).