You could study one combination that is broadly representative and is much much closer than the isolate.
The claim above about only looking at isolates was false.