And it certainly is not an argument that some method A of
ontology development is better than some other method B.
I had thought much the same with respect to XML schemas given at runtime, an
instance either conforms or not to the schema. XML Schemas could not be
classified as good or bad solely on the methodology used to create them
(although I have seen some pretty convincing arguments).
There is a larger issue though - the reconciliation and/or harmonization of
disparate models. Assuming that Ontologies may be classified as a type of
model, if six ontology developers use 2 different methodologies, I suspect
that the ones developed with like methodologies might be easier to reconcile
or map. Your point is quite correct - it doesn't necessarily make either
methodology "better" that the other since that is a highly subjective label
to bestow on any ontology development methodology.