R help - May 2005 - ANOSIM

Hello,

I am interested in using ANOSIM (analysis of similarities) which is part
of the vegan package.  When I look at the programmers notes they mention
that "They don't quite trust this method, and someone should study it
closely."  I was just wondering if this had been done, and whether it
had been found reliable or not?

Thanks,
Kerry Reeves