Applying those criterias to evolution we'll observe:
A) Evolution (i.e. macro-evolution) can't be verified. The "verifications" or evidence appealed to for macro-evolution are those of the MICRO-evolutionary kind.
Anyone who thinks there's no evidence for macroevolution obviously hasn't read Theobald's "29+ Evidences for Evolution," available at the TalkOrigins site. Here's a list of 27 different falsifiable predictions of evolutionary theory:
Prediction 1: a universal genetic code.
Potential falsification: a finding of multiple, unrelated genetic codes.
Prediction 2: a nested hierarchy of species.
Potential falsification: organisms that violate nested hierarchies, such as feathered platypuses, non-vascular plants with seeds, birds with mammary glands, insects with placentas.
Prediction 3: consilience of independent phylogenies.
Potential falsification: independently-derived phylogenies which do not converge, or which produce wildly divergent hierarchies.
Prediction 4: intermediate and transitional forms in the fossil record.
Potential falsification: fossils which do not fit into nested hierarchies, such as a mammal-like bird or an insect-like starfish.
Prediction 5: chronological order of intermediates.
Potential falsification: a negative correlation between the stratigraphy and the phylogenetic tree. E.g., mammal-reptile intermediates older than reptile-amphiban intermediates, or reptile-amphibian intermediates older than proterostome-deuterostome intermediates.
Prediction 6: anatomical vestiges.
Potential falsification: a vestigial feature that was not functional in an ancestor. Examples: snakes with vestigial wings, insects with vestigial backbones, primates with vestigial horns, mammals with vestigial gizzards.
Prediction 7: Atavisms (e.g., living whales with legs, living humans with tails)
Potential falsification: the same as the falsification for anatomical vestiges.
Prediction 8: Molecular vestiges (e.g., the broken human gene for ascorbic acid).
Potential falsification: essentially the same as for anatomical vestiges and atavisms. A finding of pseudogenes for chloroplasts in any metazoan.
Prediction 9: embryonic features of ancestors, such as gills in amniotes.
Potential falsification: embryonic features that do not exist in ancestral lines, e.g., nipples in reptile embryos or bird-like beaks in eutherian mammal embryos, leg buds in teleost fish.
Prediction 10: Present biogeography should reflect common descent.
Potential falsification: elephants on remote Pacific islands, amphibians on remote islands, Antarctic or Australian indigenous cacti.
Prediction 11: Past biogeography. We should not find the same taxon on two landmasses that separated before the taxon evolved (excepting, of course, later imports)
Potential falsification: ape fossils in South America, elephant fossils in Australia.
Prediction 12: human and ape fossils should not be found in Australia, South America, or on remote islands which would have been inaccessible to ancestral apes at the time they evolved.
Potential falsification: human,
H. erectus, Australopithicus, etc. fossils in Australia, the Americas, Antarctica, etc.
Prediction 13: Anatomical parahomology. There should be no anatomical features that are not derived from previously existing structures.
Potential falsification: an existing anatomical structure that cannot be derived from more primitive, ancestral features. A horse with wings would be a falsification, since there would be no anatomical features of any horse that could be modified into wings (no ancestors of horses have six limbs).
Prediction 14: molecular parahomology. All proteins currently in existence should statistically significant similarities to proteins with more primitive, core functions.
Potential falsification: proteins that are no related to any previously existing proteins (i.e., "new" proteins, in Dave's sense of the term, which are not derived from any previously-existing proteins). Also, derived proteins that are more deeply rooted in the phylogeny, i.e., older, than the core proteins they derive from.
Prediction 15: anatomical analogy. If two unrelated organisms evolve an analogous structure, that analogous structure must be explicable in terms of modification of ancestral structures in both organisms.
Potential falsification: gills in aquatic mammals or birds. There are no structures available in immediate ancestors from which gills can evolve. Evolution can't "skip steps."
Prediction 16: molecular analogy. If two different organisms evolve analogous molecular structures, those structures must be modifications of previously-existing structures in both organisms.
Potential falsification: no cases of molecular analogy, where all organisms that perform a function with a particular structure all use exactly the same structure; i.e., no convergent evolution.
Prediction 17: anatomical suboptimality. Since evolution can only work by modifying pre-existing structures, there should be many examples of suboptimal evolution.
Potential falsification: a mammal or reptile with no optical blind spots. Evolution cannot got back and "fix" a suboptimal design after the fact, since the ultimate use for any structure cannot be "known" by evolution.
Prediction 18: molecular suboptimality. There should be evidence of suboptimal design at the molecular level, such as the large amount of the human genome that seems to serve no known function. E.g., the human GDPH gene. There is one functional GDPH gene, and at least 20 non-functional copies of the gene.
Potential falsification: if the genomes of all organisms were efficiently designed, with only the DNA required and no more (no pseudogenes, no nonfunctional tandem repeats, "junk DNA."
Prediction 19: protein functional redundancy. There should be many genes that are common to all organisms regardless of whether they are needed. I.e., there should be genes in bacteria that also appear in humans even though they serve no function in human beings. Further, organisms which are related should have similar ubiquitous genes, and less closely-related organisms should have less closely-related ubiquitous genes.
Potential falsification: no pattern of relatedness to ubiquitous proteins. A chimp cytochrome
c protein should be no more closely related to the human version than the rat protein, or the douglas fir protein, or the yeast protein, or the
e. coli protein.
Prediction 20: DNA coding redundancy. The same pattern of relatedness should show up in the DNA coding for ubiquitous proteins. The more closely related two organisms are, the more similar their DNA sequences should be for ubiquitous genes.
Potential falsification: there should be no pattern of relatedness to DNA sequencing for ubiquitous genes, or a different pattern that is unrelated to the pattern for the amino acid sequence for the protein.
Prediction 21: transposons. Since transposons are random, but heritable, there should be a pattern in transposition that follows the phylogenetic tree.
Potential falsification: transposons that do not fit into nested hierarchies, or fit into different nested hierarchies from the phylogenetic tree.
Prediction 22: redundant pseudogenes. There should also be a pattern among pseudogenes that follows the phylogenetic tree.
Potential falsification: since pseudogenes are rare, it should be extraordinarily unlikely that the exact same pseudogene would appear in two distantly-related organisms. Therefore, pseudogenes should fit into the same nested hierarchies established by the phylogenetic tree. If there were no pattern, or a different pattern from that required by common descent, common descent would be falsified.
Prediction 23: endogenous retroviruses. Since endogenous retroviruses are heritable, their presence should mirror common descent, nested hierarchies, and the phylogenentic tree.
Potential falsification: endogenous retroriviruses which do not fit into the same nested hierarchies and patterns of common descent as the phylogenetic tree.
Prediction 24: genetic change. There should be sufficient genetic change to support the existence of macroevolution. When we compare the genomes of various organisms, we should see that genetic change traced out in the same pattern as the phylogenetic tree. Genetic change should be heritable and largely irreversible.
Potential falsification: if genomes were highly resistant to change, or commonly and typically reverted to wild type, macroevolution would be difficult to explain.
Prediction 25: the fossil record. Macroevolution predicts that as one looks at older and older sediments, one should see organisms that have increasingly primitive (in the cladistics sense of the term) features.
Potential falsification: essentially modern organisms all the way back in the fossil record, or alternatively, no pattern of primitive and derived characteristics chronologically.
Prediction 26: speciation. If speciation is an ongoing process, we should see various degrees of speciation, from fully-interbreeding populations to partially interbreeding populations to populations with reduced fertility or complete infertility to completely genetically isolated populations.
Potential falsification: if all species were genetically reproductively isolated and there were no instances of hybrids, it would be difficult for macroevolution to be true.
Prediction 27: speciation rates. Current estimates, based on the fossil record and mutation rates, are ~3 million years for complete reproductive isolation, on average. Rates of speciation and of morphological change should be as high as or higher than that observed in the fossil record.
Potential falsification: rates of morphological change that are much slower than that observed in the fossil record.