[Taxacom] shared derived molecular similarities

[Sergio Vargas]

Hi,

>I have a personal preference for the morphology being more likely correct simply because the latter can be evaluated in terms of shared derived similarities
>  for each character individually...

I have a question regarding the statement above. I'm kind of lost here... Why exactly is that you cannot evaluate individual molecular characters and establish
shared derived (molecular) similarities? say you have protein coding genes and no gaps and you can compare the sequence/structure of the proteins, etc. and
use an outgroup criterion to find these shared derived similarities, I cannot see why this is not possible...

tnx in advance

sergio