Relativity can be tested in many ways and it has (and it's suceeded). Evolution is like goo compared to relativity. Sure the evidence "verifies" evolution, except it's changed so much. It is historical in nature and not directly observed. Relativity is empirical in nature and not directly observed. Sorry if I'm having trouble defining things, but there is a difference.
Relativity has, indeed, survived the scrutiny of experiment and passed with flying colors. It is difficult to quantify "how well" a theory has been verified, but it is difficult to make a case that relativity has an advantage over evolution in this area. Certainly there has been much more published research in evolution (because the subject of evolution is a much more complex one than the subject of relativity). How often has evolution survived a falsification test - research that could have, in principle, falsified the notion of evolution? Certainly, with the discovery of DNA it could have been found that organisms thought to be related shared little or no hereditary material. Every time a new fossil is lifted from the ground, evolution stands to be falsified... what if a new mammal fossil appears in the fossil record older than the oldest amphibians, the order from which mammals derive? Once might be a mistake in dating, but enough fossils are discovered each year that a pattern of such might be established if evolution were not true. But the fact is that of the thousands of fossils that paleontologists dig up each year, and over the past 150 years, there hasn't been a single unambiguous case of a fossil organism predating the oldest of its ancestors. To avoid making the post too long, I will stop without listing any of the many, many individual results from the lab and field that independently confirm the theory in part or in whole. I am touching on some of those in the other thread...
There is a difference between historical science and non-historical science. Both are empirical (that is not the difference). The difference lies is what kind of evidence and what kind of testing is available to each. The nature of the evidence for historical events sometimes makes it too difficult to identify enough evidence to put together strong tests of a theory. Such is not the case for evolution, though.
Speciation does not guarantee "large scale evolution" over time...You assume that there is a large time scale to work with as well.
I think Rufus addressed this point well. Natural selection, changing environments, and speciation do guarantee large scale changes over large periods of time. You cannot split a collection of similar objects into groups, make different changes to each group, split the sub-groups, repeat again, and continue doing this for very long before the individual groups have ceased to bear the slightest resemblance to one another. That is large scale evolution. It does not necessarily guarantee the evolution of
novelty, though. What does guarantee the evolution of novelty (complex adaptations like eyes and wings), is variability with a feedback mechanism. To see this illustrated, play one of those games of cellular automata. Genetic algorithms have been used to allow computers to program themselves by sucha feedback loop. They have "evolved" very novel systems (ones that engineers are challenged to even understand the workings of) that do remarkable complex tasks. For instance, a computer coupled to a box of components and re-wirable circuitry programmed itself to be a radio using genetic algorithms (a computer simulation of variation plus natural selection). A computer programmed itself to improvise jazz solos using the same methods. Both in very short periods of time, thanks to a strong selective pressure.
Nice. So the 3 coccyx vertebrae are enlarged and not fused? It looks like a tail, but perhaps it is merely a pathology? The "5 vertebrae tail" (figured 2.2.1. talkorigins) is not elaborated on. Did that one have extra coccyx vertebrae, or were the S5 and S4 vertebrae part of the "tail"? This is interesting and provides something to talk about, but it doesn't prove anything to me.
The one pictured in the x-ray only had 3 "tail" vertebrae, and they were homologous to the three small "fused" vertebrae in the coccyx. The vertebrae labeled S1-S5 are the sacral (meaning spinal) vertebrae.
We are working from the same material here, so I cannot elaborate much more than the author without going to the library. He does cite his references well for those of us who require a more in depth examination. The most important part, to me, is that the true human tail is not merely a deformity, as witnessed by:
The true human tail is characterized by a complex arrangement of adipose and connective tissue, central bundles of longitudinally arranged striated muscle in the core, blood vessels, nerve fibres, nerve ganglion cells, and specialized pressure sensing nerve organs (Vater-Pacini corpuscles). It is covered by normal skin, replete with hair follicles, sweat glands, and sebaceous glands (Dao and Netsky 1984; Dubrow et al. 1988; Spiegelmann et al. 1985). True human tails range in length from about one inch to over 5 inches long (on a newborn baby), and they can move and contract (Baruchin et al. 1983; Dao and Netsky 1984; Lundberg et al. 1962).
(from the same page.) The references are cited, and where possible links are given in the footnotes to the abstract on PubMed.