Scientific method used in studies of co-evolution

Believe it or not, I actually used the scientific method in my studies of co-evolution. First I stated a theorem of co-evolution. Then I theorized that co-evolution would be much more difficult in interspecies relationships where: (1) there is totally co-dependent mutualism as opposed to other kinds of interspecies relationships, e.g., predator-prey, parasite-host; (2) the required changes in both organisms are of kind and not just degree, i.e., the changes cannot be completely gradual; and (3) the traits are totally dependent (i.e., fatal or very harmful in the absence of the corresponding trait in the other organism) in both or at least one of the organisms. I then searched for such interspecies relationships and I think that I found a fairly good example in buzz pollination.

One scenario in which the evolution of buzz pollination might occur is that some flowers with regular exposed pollen developed tube-contained pollen while retaining the exposed pollen and that the exposed pollen feature then disappeared after buzz-pollinating insects appeared. But this is a very complicated evolutionary pathway for unguided evolution.

Also, as I said, a basic problem for co-evolution of co-dependent mutualism is that in order to have a benefit, corresponding changes in both organisms must co-exist at the same time and the same place, and that is unlikely when changes are rare (and is especially unlikely when the changes are fatal or very harmful in the absence of corresponding changes in the other organism). This is a problem even for gradual changes.