ptman writes:
If you did have a complete set of intermediates, you would only be able to define arbitrary breakpoints between them and the sum would be Macro but the individual changes would be Micro.
Your thinking is rather linear and evolution is not.
Think in terms of a heavily pruned tree with only a few branches extending to give rise to others.
The branch points are speciation events.
Looking at terminal branches on different sides of the tree (extant species in disparate orders, for example) they can all be traced back to a common branch deep in the tree, a branch that now represents the divervence of some higher order taxon, even though it was originally only a speciation event.
The tree represents macroevolutionary structure, or phylogeny.
This is the only useful application of the concept.
The tree structure doesn't really hinge in any way on the presence or absence of 'intermediates' - remember these are only defined a posteriori as forms that apparently fit well between an earlier form and a later form.
They are not in any way essential to make sense of higher evolutionary processes. We now have far more powerful molecular tools to infer the structure of the tree than simply searching for fossil intermediates.
And the branch points, or 'breakpoints' as you call them, are not in any way arbitrary. That is incorrect. They are inferred speciation events that *had* to occur at certain points in order for taxa to diverge, unless you reject outright the idea of commonality of descent. Just because we may not know the exact details about how and when particular speciation events occurred in phylogenetic history, this does not make inferences of their occurrence in any way arbitary.