Anything is "all rubish" if you never actually read the work of the
people working on the "rubish." I am amazed that you would contend that
variation in clock speed is ignored by those working in molecular
systematics. The algorithms are constantly scrutinized and improved, and
part of that process is, consistently, assessing their performance under
conditions of gross differences in rates of molecular evolution. There
are other factors also investigated, such as compositional bias. Far
from feeding an algorithm and accepting the answer, anyone who publishes
in a reputable journal should pass the scrutiny of reviewers who insist
that the potential pitfalls be addressed (of course, as in any science,
some papers do get published without appropriate scrutiny).
