Even though you don't give a proper attribution, this idea seems to be
based on research by Christian Schwabe, who did indeed conclude on the
basis of this and a small number of similar results that the
evolutionary tree needs some adjustment. (Schwabe, C., Büllesbach, E.
E. (1994) Relaxin: structures, functions, promises, and nonevolution.
FASEB J. 8,1152-1160) More specifically, he argues that chemical and
biochemical principles drive evolution more than genetic mutation, and
that life on Earth has a multiplicity of origins from chemical pools
of nucleic acids.
For the purpose of your post, we can leave aside if this
interpretation would follow assuming his empirical data were correct,
and simply focus on the data. The article is from the 1990s- in
science it is always necessary to check if results were confirmed by
later research, and in this case, it seems this is not the case. His
research relies heavily on PCR and microsequencing data, which is
known to be in danger of contamination. In this case, it seems that
later research came to different results, see Wilkinson, T. N.,
Speed, T. P., Tregear, G. W., Bathgate, R. A. (2005) Evolution of the
relaxin-like peptide family. BMC Evol. Biol. 5,14
One should note that even if his results had been confirmed, the
general pattern of similarity would have been unaffected, it is
extremely strong. There are a few "anomalies" - biology is not quite
as tidy, and driven by universal laws, as physics. But typically, it
is quite straightforward to find a reason why that anomaly occurred,
e.g. due to paralogy (there is more than one relaxin gene, have the
correct counterarts been studied?) or
the long branch effects, and a couple of others more.