Bullshit. There are plenty of species that breed fast enough, in large enough numbers, that we can observe huge numbers of generations of them in a controlled environment. Typically insects are used, simply because they are easier to observe than bacteria, but we actually observe it in bacteria *all the time* in the wild; drug-resistant strains, or new variants of things like bird flu, or so on. It's also been observed in the lab, though; take a colony of insects (like fruit flies), separate them, apply an environmental pressure that selects for different traits, wait, observe adaptations. Eventually the adaptations lead to new, non-interfertile species.
It's those final points - the arising of adaptations and their spread throughout the population, and the speciation - that are the directly testable aspects of evolution by natural selection. There are others, of course; things like the ability to predict the presence of species bearing particular characteristics at certain chronological points in the fossil record (which we have later found), the presence of vestigial organs which are no longer used or needed but were in evolutionary ancestors, the parallel nature of the morphological classification trees (classifying life based on observable characteristics) and genetic classification trees (based on DNA), which was predicted, and found to be true, even when the species would have needed to diverge hundreds of millennia ago... So many things.
I think your problem is that you don't understand what a prediction means in this case. Genetic modifications to produce specific changes has nothing to do with evolution. That's just testing genetic theory in general, which nobody seems to have any problem with. Evolution doesn't predict the specific random mutation that will occur, and nobody who knows anything about it would suggest it does. In fact, the random nature of evolution by natural selection means that two different populations, both subjected to the same environmental stress, may adapt *differently* to it. This is observable even in human populations (which are only a few hundred thousands of years old), where different groups of humans that adapted to high-altitude environments (low oxygen) did so in different ways.
What the "theory of evolution by natural selection" predicts is that there *will* be mutations (directly observable and tested), that a mutation (not *the best mutation" because all it needs to be is good enough to increase the chance of successful reproduction) which better adapts the organism to the environment will become widespread within a population (also directly observed in the lab), that this process will continue over time as environments change and new, superior adaptations occur in the usual course of random mutation (observable from the fossil record, with predictions about what will be found in the "holes" in the fossil record also found to be accurate), and that over time this leads to speciation (also observed in the lab).
It is in every way a scientific theory, and one that has had its predictions tested and verified time and again. Comparing it to an Earth-centric universe model just further shows that you don't understand science at all; there was plenty of evidence which didn't fit that model, and the predictions that would have arisen from it (for example, that "anything which orbits the earth at different speeds, such as the sun and the planet Venus do, must at some point in their orbits come to be on opposite sides of the Earth") are easily observed, even with the instruments of the day, to be false. No scientist actually concerned with accuracy, rather than the importance of appearing to be accurate, would have seriously defended such a theory. In fact, this is much the same situation as concerns creationism today; people want to believe that they're important and special, and they have an old book that tells them so, and they therefore oppose anything which appears to conflict with the aforementioned book for the sake of upholding the appearance that the book is right (and therefore that they are right about their superiority in the other ways the book mentions) rather than concerning themselves with any form of scientific evidence.