It doesn't. We still have bacteria, for instance. Once an organism breaks through the fitness constraints into the next space, the equation is no longer A<=B but, instead, is A'<=B' or something else.
Can you give an example of an observed instance (fossiles arn't observed instances ;)?
There are numerous paralogous and homologous structures in extant species which provide examples.
You also see, I trust, that this largely explains the discontinuity between species.