Two accounts of one plant
Genomic, chemical and phenotypic work each produce a description of a plant, and none of them is the whole description. Put beside decades of breeding observation, they either agree or they do not, and both outcomes are informative.
Evidence, not verdict
Characterization is best understood as accumulating evidence rather than as settling a question. Identity, relationship and distinctiveness are all easier to argue for with several independent lines of information than with one.
It is also worth being precise about which stage a piece of work is at. Research underway, a capability being developed and a finding that has been demonstrated are three different claims, and collapsing them is how a research programme loses credibility.
The useful question is not which account is right, but what it means when two independent accounts disagree.
