I am specifically mentioning hummingbirds, since that's the title, in the image, and the bulk of the discussion in this thread when I looked at it this morning. That being said, no, I don't think it has to be hummingbirds.

Keep in mind, these flowers are significantly smaller than most (if not all) hummingbirds, and hummingbirds themselves are amongst the smallest, if not the smallest, birds in the world. So, any species of bird that these flowers attract are going to be significantly larger than the flowers.

Not to mention that virtually all birds that feed from flowers have excellent visual acuity, that's almost a requirement for them to be successful with this feeding strategy. It seems highly unlikely that the birds would have a difficult time identifying that these are flowers and instead confusing them as a flock of miniature birds feeding from the plant. And, to me, that pretty much negates any argument for selective pressures on the flowers to have birdlike appearance for that purpose.

At least with the example of the "bee orchids" I mentioned in a prior comment, the selective pressures and the overall context make sense and appear to explain why the flowers have evolved to look like female bees and wasps. But the "hummingbird flower" from Australia that kind of resembles the profile of a hummingbird if viewed from a specific angle and out of context doesn't really hold up, in my opinion.