[From many - one.]

Here's a sample of eyespots (scroll down):
http://linus.socs.uts.edu.au/~don/larvae/satu/helen.html

Here's a moth that looks and acts like a hummingbird, enough to fool professional naturalists in the field albeit briefly:

Hummingbird Moth sipping nectar from a Bee Balm
http://www.skipper-systems.com/pages/hmoth.htm

[Liberal Classic]

I've seen those Hummingbird moths. They fooled me, too. I remember thinking, "look at those hummingbirds feeding from the azeleas." But when I moved in for a closer, it was quite a surprise. They moved just liked hummingbirds, very rapid wingbeats, very fast zooming motion. They were hard to focus on for any length of time. It wasn't until I glimpsed on at a flower that I realized I was looking at a big bug.

[Ichneumon]

Here's a moth that looks and acts like a hummingbird, enough to fool professional naturalists in the field albeit briefly

There's also the bumblebee moth, which looks remarkably like a bumblebee, especially when in flight:

[AntiGuv]

OK, thanks! That's all I needed to see.

1) Those eyespots hardly resemble a bird's face. They are scare patterns on the butterfly wings. In fact, the last thing the moth wants to resemble is a bird - that would attract birds, not keep them away. What they want to resemble is a predator of some kind, to frighten birds away.

Hence my confusion.

And so to answer the original question, butterfly eyespots are hardly "irreducibly simple" (or "irreducibly complex" or whatever floats your boat). There are any number of rudimentary patternings short of nice-neat eyespots that confer a survival advantage.

You can view a variety of such patterns here. (That was a 5 second Google search BTW). It's hardly difficult to imagine how less precise camouflage patterns or 'scare patterns' might in a few species hone themselves into very precise facsimiles, but in others remain more imprecise (yet still beneficial enough for their purpose).

2) A moth that resembles a hummingbird is hardly surprising. The moth has simply evolved convergent features for convergent purposes. It's not as if the moth was trying to look like a hummingbird. The moth was presumably trying to be a creature that hovers and sips nectar from deep flowers, and that just so happens in this case to resemble a hummingbird because that's what hummingbirds are perfecting as well.

[AntiGuv]

PS. IIRC Moth eyespots usually appear to mimic cat eyes, so that when the moth is in foliage the bird will discern just the 'eyes' and think there's a cat prowling around.