You can get lots of variation and lots of new species from mere splits in the population, but eventually enough evolution in one direction will lead to the condition of depleted genetic diversity from which further evolution is impossible. Evolution defeats evolution. You can never get anything but variations on a species, never a truly new species.
If your description of evolution was correct... then your conclusion would also be correct.
However, your description of evolution is not correct. It's not really "wrong" though... just more "incomplete."
Your description of evolution does accurately (enough) describe a very small subset of evolution and how some very small specifics actually work. And if that was the only way evolution worked... you would be correct and your conclusion would, again, also be correct.
But it is, in fact, only a very small subset. There are many, many other ways evolution works. Most of those other ways include an increase in genetic information (from mutations) along the way.
In fact, even the specific subset you describe includes mutations and a resulting increase in genetic information. It's just that the specific subset you describe 'evolves' faster than the mutation rate generally has to increase the genetic information by any significant amount. And you still end up with the 'net loss of genetic information' in the small area you're describing.