Your figure illustrates the broad phylogenetic branch points and isn't
fine enough in detail to get at your request. The figure isn't to
scale in terms of time. Humans and chimps had a common ancestor just
around 5 million years ago and feathers evolved well over 100 million
years ago. The accuracy of the figure may be in doubt (Where they
have placenta under mammals they likely should have eutherian because
both eutherian and marsupials have placenta. It looks like your
figure is just trying to show the major points of what we call a
nested heirarchy. You have traits nested within traits due to descent
with modifications. You have vertebrates (including fish to
mammals). You have terrestrial vertebrates from amphibians to
mammals. You have amniote terrestrial vertebrates (reptiles, birds
and mammals). The branch point of crocodiles and birds from reptiles
and their relationship with mammsls seems to be messed up on this
figure. Mammals are more closely related to crocs and birds than to
snakes and lizards.
For something like the chimp and human common ancestor it has to be
inferred from the genetic evidence. Just as there is a nesting of
traits shown in your figure there is a nesting of genetic changes over
time. We can tell that a fish genome evolved from a more primative
cordate genome. The amphibian genome evolved from a fish genome. The
terrestrial amniote vertebrate genome evolved from an amphibian
genome. The egg laying mammalian genome evolved from something that
most people would call a reptile, but the purists call it a synapsid.
The eutherian and marsupial mammal genomes evolved from an egg laying
mammalian genome. Primate genomes evolved from a eutherian genome and
chimps and humans evolved from a primate genome. You can tell this by
the nested similarities and differences.
Just as we know that second cousins are related to a great grand
parent that they have in common by genetic testing, even if we don't
have the great grand parent's DNA, we can tell that chimps and humans
shared a common ancestor using the DNA evidence. We don't have to
know who that great grand parent was just that the DNA evidence says
that they existed.
To try to identify the common ancestor using fossils is pretty
futile. There are so many closely related species existing at any one
time that when you find a fossil that looks right for the common
ancestor you can't really tell even if it falls at the expected point
in time. Just look at how many deer species there are around the
world and then project out 5 million years into the future when most
of those species are extinct and try to figure out which deer like
species was the common ancestor of what survived. All you will
basically know is that the common ancestor looked something like a
deer.
You can identify the ancestor of more recent species. The easiest
example are the amphidiploids (mostly plants). These are
allotetraploids that are new species due to the hybridization of two
existing species and the tetraploidization of the resulting hybrid
product. When a hybrid is created between two different species it is
usually sterile, but by a quirk in mitosis the chromosome number can
double and now there are good pairs of each chromosome from each
parent species. The resulting species is genetically isloated from
both parents and has many more chromosomes than either parent
species. There are examples of this that have been identified in the
field where the parental species still exist. So you essentially have
a branch point where you have the original species coexisting with the
new species. All vertebrates from fish to mammals may have evolved
from a common ancestor that evolved either as a tetraploid or an
allotetraploid. All vertebrates share that original doubled genome.
There are other methods of speciation, but that is one of the easiest
to observe and it has had a big impact on vertebrate evolution, and we
can identify the species that the new species evolved from.
Ron Okimoto