There is simply no evidence that speciation occurs as the result
of a single point mutation, or even that it occurs in a single
generation.

As a general rule, except for fatal mutations and mutations causing
sterility, individuals with mutations breed quite successfully with
other members of the same species.  This leads to a certain mixing
of the genes, and to the presence of a number of individuals with
any given variant. (The number depends somewhat on the reproductive
potential of the mutant).  Thus mutation, per se, is unrelated to
speciation, it just increases the amount of variability in a given
population.   Just that, nothing more.

The currently favored model of speciation is the one developed by
Ernst Mayr in the 1940's, often called the peripatric model.  The
key event in this model is the establishment of a small, isolated
population in a marginal habitat for the prospective parent species.
Several effects are enhanced, or enabled, by this situation.  They
are genetic drift, differential selection, and release from pressure
to conform.  In combination these factors can cause *relatively* rapid
divergence in these isolated populations [that is divergence in
periods of only a few hundred generations].

This level of divergence can easily lead to reproductive barriers
between the local isolate and the original species.  When this point
is reached a new species is said to have formed. (although the exact
degree of isolation necessary to qualify as a species is uncertain -
and almost all intermediate degrees of reproductive isolation are
known from real populations).

Other modes of speciation do occur, but are generally held to be
rare, and thus unimportant in the long run.