So your boundaries are not natural ones and should not be extrapolated to include potential variability in the wild
But there are natural boundaries in the wild. Boundaries due to geography for example.
Every step in the transformation of one kind into another would have to have survival advantages
Yes. In a given environment. However the environment is also changing.
That is what the fossil record actually shows, natural selection appears to keep things within limits
Yes. Natural selection can only build upon existing material in small incremental steps.
Some things exhibit stasis and others become extinct.
No. Some things exist in an environment that has not changed significantly for a long time. These are basically static. Some things adapt to the changing environment. These we see transition accordingly. Other things fail to adequately adapt and become extinct. This is what we see.
We have no scientific reason to believe that any one species became another species just because they may have certain features in common
Really? What about cases where common sense similarities are almost non-existant but the chronological fossil record and genome studies are in direct agreement?
Who would have thought that a whale and a cow were genetically more closely related than a whale and a seal, for example?
On what basis could intelligent design possibly predict this?